Journal of the Scientific Agricultural Society of Finland Vol. 48: 203-304, 1976 Maataloustieteellinen Aikakauskirj a TRANSMISSION OF COCOA VIRUSES BY MEALYBUGS (Homoptera: Pseudococcidae) Selostus: Villakirvat kaakaopuun virusten siirtäjinä OSMO ROIVAINEN 1) Cocoa Research Institute of Ghana, P. O. Box 8, Tafo, Ghana l) now in Agricultural Research Centre, Department of Pest Investigation, 01301 Vantaa 30, Finland To BE PRESENTED, WITH THE PERMISSION OF THE Section of Mathematics and Natural Sciences of the Philosophical Faculty of the Univer- sity of Helsinki, for public criticism in Lecture Room of Deparment of Zoology, Pohjoinen Rautatiekatu 13, on 20th August 1976, at 12 o’clock noon SUOMEN MAATALOUSTIETEELLINEN SEURA HELSINKI https://www.c-info.fi/en/info/?token=q7RK7m7Rp7_GpEqr.gJFCMH1iTh-z47ELI6afMQ.FbWm0rTPU4m1g-OVEskA_x-HoIAO5FW0HOm7cDfpo6N_s__8d1vmwK_gd4GPxBr-UlTOHZET_x01jBV1jMOB7HKGIjvvFCRFcg6OiO9L-jeHytdj57irT_5iJU1SrmmGxmcItpds6ZQE1eKfCptSpQeF_efqy3euP_uKwA TO ELSIE Preface In the course of preparing this study I have received help from many persons. Since this work started early in 1966 it is possible that some valuable assistance may have been overlooked. Dr. R. H. Kenten, Dr. J. T. Legg and Dr. G. K. Owusu gave very useful criticism at various stages of progress of the work. Advice on statistical matters was given by Mr. N. A. Goodchild, Dr. S. C. Pearse and Mr. P. Walker. Prof. Y. Seppälä helped in some mathematical questions. I note with pleasure the support given by Prof. L. Cutcomp, Dr. Leo E. LaChance, Dr. D. A. Lindquist and Mr. O. E. S. Lloyd of the International Atomic Energy Agency. At the early stages of this work it was very useful to meet Prof. A. F. Posnette and get his expert advice on various aspects of the work and suggestions on useful lines of investigation. Mr. E. K. Tetteh was very helpful in matters concerning the literature. My supporting staff, and others directly involved in the execution of the various ex- periments, statistical calculations and other aspects of the work, were the following; Mr. E. O. Agyare, Mr. B. S. Asare, Mr. M. E. Bonku, Mr. J. K. Bonney, late Mr. E. M. Ewool, Mrs. M. Ewool, Mr. E. M. Ewool Jr., Miss A. Vasarainen and Mr. Y. E. K. Yirenkyi. All these persons gave invaluable assistance in the cource of this study. My teacher in zoology, Prof. E. Paimen, has encouraged me in many ways and followed the progress of this work with keen interest. The language of this paper was inspected by Mrs. L. Ritarsalo. I gratefully acknowledge the help of all these people and wish to express my sincere gratitude to them. Permission to publish this paper has been obtained from Dr. E. J. A. Asomaning, Director of the Cocoa Research Institute, Tafo, Ghana, and the International Atomic Energy Agency, Vienna, Austria. I am grateful to the Scientific Agricultural Society of Finland for accepting this study into the journal of the society. Uppsala. 26 November 1975 Osmo Roivainen CONTENTS page Abstract 209 1. INTRODUCTION 209 2. EXPERIMENTAL TECHNIQUES, CONDITIONS AND MATERIALS 211 2.1 Collection and handling of mealybugs 211 2.2 Attempts to rear P. njalensis 212 2.3 Artificial feeding 214 2.4 Feeding cones and cages 214 2.5 Physical conditions 215 2.5.1 Reactions of mealybugs to light 215 2.5.2 Reactions of mealybugs to humidity 217 2.5.3 Reactions of mealybugs to temperature 218 2.5.4 Discussion and conclusions 220 2.6 Temperature and reproduction 221 2.7 Use of radioisotopes 223 2.8 Biological half-life of S2 P in mealybugs 224 2.9 Virus source plants 225 2.10 Virus test plants 227 2.11 Moulting cages 227 2.12 The standard virus transmission experiment 227 2.13 Statistical calculations 229 3. EFFECT OF VARIOUS FACTORS ON VIRUS TRANSMISSION 229 3.1 Relationship between food uptake and infection rate 231 3.2 Relationship between plant and mealybug radioactivity 232 3.3 Mealybug honeydew and plant radioactivity 232 3.4 Radiation effects 235 3.5 Preliminary fasting/feeding 236 3.6 Availability of virus in source plants 240 3.7 Acquisition feed 242 3.7.1 Removal of mealybugs from virus source plant 242 3.7.2 Length of acquisition feed with CSSV 1A and P. njalensis 244 3.7.3 Length of acquisition feed with other isolates and P. njalensis 246 3.7.4 Length of acquisition feed with CSSV 1A and F. virgata 248 3.7.5 Acquisition of virus from liquid leaf extract 248 3.8 Persistence of virus in mealybugs after acquisition feed 251 3.8.1 Persistence of CSSV 1A in fasting P. njalensis 251 3.8.2 Persistence of CSSV 1A in feeding P. njalensis 251 3.8.3 Consecutive transmission of virus by a single mealybug 255 3.8,4 Persistence of CMLV 1C in fasting P. njalensis 255 3.8.5 Persistence of CSSV 1A in fasting F, virgata 256 3.8.6 Relationship between virus persistence and time 257 3.8.7 Persistence of virus through postacquisition moulting 258 3.9 Inoculation feed 259 3.9.1 Latent period of virus in the test plant 260 3.9.2 Number of mealybugs on test plants during inoculation feed 261 3.9.3 Inoculation feed on different cocoa types as test plants 261 3.9.4 Location of feeding on test beans 264 3.10 Effect of physical factors on virus transmission 265 3.10.1 Effect of light on virus transmission 266 3.10.2 Effect of temperature on virus transmission 267 3.11 Relationship between virus uptake, persistence and infectivity of mealybugs 267 4. DISCUSSION AND CONCLUSIONS 272 4.1 General 272 4.2 Preliminary fasting/feeding 273 4.3 Availability of virus in source plants 274 4.3.1 Availability of virus in different cocoa types 276 4.4 Acquisition feed 277 4.4.1 General conditions during acquisition feed , 277 4.4.2 Length of acquisition feed 278 4.4.3 Acquisition of virus from liquid leaf extract 280 4.5 Persistence of virus in mealybugs after acquisition feed 281 4.5.1 Persistence of virus in fasting/feeding mealybugs 281 4.5.2 Persistence through postacquisition moulting 282 4.6 Inoculation feed 283 4.6.1 Length of inoculation feed 284 4.6.2 Number of mealybugs on test plants during inoculation feed 284 4.6.3 Inoculation feed with virus resistant/tolerant cocoa 285 4.7 Effect of vector age on virus transmission 286 4.8 Efficiency and virus/vector specificity of vector species 288 4.9 Effect of physical factors on virus transmission 289 4.10 Transmission of virus complexes 290 4.11 Latent period of virus in the vector 291 4.12 Type of transmission 291 SUMMARY 293 REFERENCES 295 SELOSTUS 300 APPENDIX I VECTORS OF COCOA VIRUSES 301 APPENDIX II VECTORS OF COCOA VIRUS ISOLATES 302 APPENDIX 111 INSECTS THAT HAVE FAILED TO TRANSMIT ANY OF THE COCOA VIRUSES 304 209 JOURNAL OF THE SCIENTIFIC AGRICULTURAL SOCIETY OF FINLAND Maataloustieteellinen Aikakauskirja Voi. 48: 203—304, 1976 Roivainen, O. 1976. Transmission of cocoa viruses by mealybugs (Homoptera Pseudococcidae). J. Scient. Agric. Soc. Finl. 48; 203—304. Abstract. Several mealybug species reacted positively towards light and high humid- ity. Planococcoides njalensis (Laing) had a thermopreferendum at 20 25° C. and 25 33° C. was optimum for reproduction with a partial fecundity of 92 nymphs per female. T J biol. of 32P in the species was 11.4—15.3 days. Food uptake from virus infected plants was positively related to the infectivity of P. njalensis with cocoa swollen shoot virus isolate IA. Plant and mealybug radioactivity were positively correlated. Amount of honeydew excreted by the species and the radioactivity of the honeydew were positively correlated. Frequency distribution of 32P in P. njalensis after feeding on 32P-active seedlings or sucrose solution was similar. Radiation effects on infectivity of mealybugs were not detected at 30 //Ci/ml of 32 P in seedling culture solution. With CSSV 1A and P. njalensis, preliminary fasting increased the rate of settling down to feed, the food uptake and infectivity of mealybugs. Virus was transmitted more often from stems than leaves of source plants. Groups of seedlings of 29 —54 days old were almost equally good sources of virus for mealybugs and availability varied more between individual source plants. Mealybugs removed their stylets from source plants in 240 seconds. The length of acquisition access time for maximum transmission was 48—72 hours. Virus was not acquired during short feeds of up to one hour. Virus was acquired from infected seedlings by mealybugs via a parafilm membrane. Acquisition of virus from liquid leaf extract was rare. Virus persisted in mealybugs up to 72 —96 hours and loss of infectivity was exponentialwith aTJof 12—l3 hours. Virus persisted in the mealybugs through postacquisition moulting. Mealybugs vere seemingly more efficient vectors after postacquisition feeding than fasting, Latent period of virus in the test plants was short and over 98 % of the test plants showed virus symptoms in three months. With different numbers of mealybugs on test plants during the inoculation feed, the infection rate was as expected. The use of mealybugs while testing virus resistant and tolerant cocoa is demonstrated. Location of feeding mealybugs on test beans did not influence the infection rate. Virus transmission by mealybugs was not affected by light or darkness. The food uptake was highest and virus was transmitted most often in 29—36° C. The transmission of CSSV isolate 1M or cocoa mottle leaf virus isolate 1C with P. njalensis as the vector, or CSSV 1A with Ferrisia virgata (Cockerell) was similar to that of CSSV 1A with P. njalensis as the vector. A formula is given to describe the increase of infectivity in mealybugs during acquisition feed. 1. Introduction Our knowledge of viruses transmitted by mealybugs (Pseudococcidae) is almost entirely based on the extensive research carried out with cocoa swollen shoot disease, a scourge in the cultivation of the cocoa plant (Theobroma cacao 210 L.) and one of the most economically damaging of all plant virus diseases. The disease was first discovered in Ghana by Steven (1936), and Posnette (1940) showed through successful graft transmission that it was caused by a virus. Posnette (1947 a) also gave a detailed description of the virus. The search for vectors of cocoa swollen shoot virus (CSSV) started in about 1940 (Posnette 1941, Cotterell 1943), and Box (1945) as well as Posnette and Strickland (1948) confirmed that mealybugs were the vectors of this virus. Subsequently, other mealybug transmitted cocoa viruses have been found, and cocoa viruses occur also in Nigeria, Ivory Coast, Togo, Sierra Leone, Trinidad and Sri Lanka. Some aspects of cocoa viruses were reviewed by Thresh (1958 a, b) and Thresh and Tinsley (1959). A general account was prepared by Dale (1962). Roivainen (1973) has reviewed the ecology of cocoa viruses and their mealybug vectors. The most up-to-date general review on cocoa viruses was recently made by Thorold (1975). Cocoa viruses and isolates obtained from different localities differ usually in symptoms and may also differ in host range, virulence and characteristics of mealybug transmission. There is little information on the relationships between West African isolates and those from other countries, and even the relationships between the various West African isolates themselves are not entirely clear. Thresh and Tinsley (1959) classify the West African mealybug transmitted isolates into two groups, cocoa swollen shoot virus (CSSV) and cocoa mottle leaf virus (CMLV), on basis of symptoms, host range and vector specificity. Kenten and Legg (1967, 1971), however, have shown that these two groups of viruses have many similarities and are serologically related, and this suggests that both groups should be referred to as CSSV (*/*; */*:U/*:S/Cc). The status of cocoa Trinidad virus (CTV) is not clear as serological tests have not been done, but some dissimilarities between CTV and the West African viruses suggest that these are different viruses (Posnette 1944, Baker and Dale 1947, Kirkpatrick 1950). For the sake of convenience CMLV, CSSV and CTV are treated as distinct viruses in this paper. Little is known about the Ceylon cocoa virus (CCV) which is also transmitted by mealy- bugs (Carter 1956). Two other distinct cocca viruses are known, the cocoa yellow mosaic virus (CYMV) and the cocoa necrosis virus (CNV), but neither are mealybug-borne (Posnette 1950, Attafuah and Brunt 1960, Blencowe et al. 1963, Owusu 1971 a). Several species of mealybugs are known to transmit one or more of the four cocoa viruses or their different isolates (see Appendix I and II). Much infor- mation on characteristics of transmission comes from the numerous contribu- tions dealing with Planococcoid.es njalensis (Laing). Some other species, e.g. Ferrisia virgata (Cockerell), Planococcus citri (Risso), and Dysmicoccus brevipes (Cockerell), have also been studied but to a lesser extent, and little is known about the transmission characteristics of the other vector species. Among the most important transmission characteristics studied by many authors and reviewed by Dale (1962), Roivainen (1973), and Thorold (1975), are that mealybugs usually require several hours of acquisition feed on infected plants to become infective, but can then infect a healthy plant in less than an hour. 211 The transmitting efficiency of P. njalensis increases with prolonged feeding on infected plants and maximum efficiency is obtained after an acquisition feed of 16 hours or more. The infectivity of vectors is usually lost soon and not more than two plants have been infected consecutively by the same mealy- bug. In terms of a maximum, however, the virus has persisted up to about two days in starving mealybugs. Virus can occasionally be acquired and trans- mitted by mealybugs within five hours. The virus generally affects mainly conducting tissue (Knight and Tinsley 1958) and, according to more recent investigations by Entwistle and Longworth (1963), it is obvious that mealy- bugs acquire virus from the phloem and must deposit virus in the phloem to infect a plant. In general, viruses with intermediate transmission characteristics are called semi-persistent in a system which designates transmission as persistent (circula- tive) or non-persistent (stylet-borne) (Sylvester 1958). In such a system the mealybug transmitted cocoa viruses best fall into the group of semi-persistent viruses. As to the mechanism of transmission, evidence has been presented to support the view that some semi-persistent viruses are stylet-borne, although there is no general agreement that all the viruses of this group are stylet-borne. Recently Roivainen (1971) has reported successful persistence of CSSV through a moult in P. njalensis, which suggests circulative transmission mechanism of this virus in the mealybug vectors. The experimental work of the present study was carried out in 1966—71 at the Cocoa Research Institute of Ghana, and some interim reports on the progress of this work have been made available earlier (Roivainen 1968, 1969, 1970, 1971, 1972 a, b). The aim of the present paper is to give coherent and up-to-date views of the various factors which have an influence upon the rela- tionships between the viruses of the cocoa plant and their mealybug vectors. Much of the scattered information on mealybug transmission of cocoa viruses is available only in the various reports of the West African Cocoa Research Institute or, following the dissolution of the West African Research Organiza- tions in 1962, in reports of the Cocoa Research Institutes of Ghana and Nigeria, and these reports have a very restricted distribution. Part of this information is conflicting and some of it is difficult to assess because details of experimental procedures are lacking or obscure. Together with presentation of primary experimental data the purpose of this study is to collate and critically assess published information on the basic aspects of the cocoa/virus/vector relationships to provide a firm basis for further work and to indicate useful lines of investigation. The natural spread of virus by vectors or control measures are not discussed in this paper. How- ever, some relevant aspects of vector biology are included as deemed necessary. 2. Experimental techniques, conditions and materials 2.1 Collection and handling of mealybugs Most of the virus transmission tests were done with P. njalensis, but also with F. virgata to some extent. P. njalensis was collected in the field as colonies 212 on cocoa pods or young cocoa shoots. When these parts of the plant are tapped and left for a few minutes, most of the insects withdraw their stylets and can be collected easily with a fine brush without damaging them. The collection was done by a team of ’bug-hunters’ which brought in the daily catch from various cocoa plantations of the Cocoa Research Institute. Because of possible differences between the daily collections, control tests were carried out when necessary. A collection brought into the laboratory consisted of mealybugs in all stages of development, and in preliminary experiments the collection was used without refinements as to different stages or impurities of honeydew and waxy skin secretions. In most of the experiments, however, only young nymphs of known age were used. These were obtained from reproduction cages specially constructed for the purpose (Fig. 1). The cage consisted of two small chambers separated from each other by a wire or plastic net of a mesh that the adult females could not walk through. Field collected adult females were introduced into one of the two chambers and after one day or a few days newly- borne nymphs, reproduced by the females, could be collected from the other chamber. Nymphs adhering to the waxy covering of the females could be separated mechanically by tapping and shaking the cage and using the net as a sieve at the same time. F. virgala was collected from cocoa to some extent, but most often from a variety of ornamental plants. This mealybug species was handled in the reproduction cages like P. njalensis. 2.2 Attempts to rear P. njalensis Large numbers of mealybugs were often needed for various experiments, but the method of obtaining supplies by field collection had its drawbacks. The risk of collecting from virus infected trees meant that mealybugs had to Fig. 1. A two chamber repro- duction cage with filter paper between the snap-on lids and chambers. The net in between the chambers keeps adults in one chamber but nymphs can easily leave this chamber and enter the other. 213 be starved to free them from possible contamination of virus, or by the use of reproduction cages, virus free nymphs could be obtained because transovarial passage of virus does not occur (Dale 1958). In both cases some mealybugs were lost always. Furthermoie, there were periodic shortages of mealybugs during the dry season or after the harvest of cocoa pods and, therefore, supplies could not be maintained at a satisfactory level continuously. Thus a suitable method of mass-breeding mealybugs of clonal origin would have been extremely useful. Two experiments were carried out to rear P. njalensis. In the first, two wild Theobroma-species and five types of cultivated cocoa were tested. Seven replicate seedlings of each species and type were grown in plastic buckets and later infested with mealybugs and attending ants, Crematogasler slrialula Emery. Mealybug numbers were assessed about every four weeks by counting the live adult females on the seedlings. The mealybugs had obvious difficulties in becoming established on T. grandiflora and T. obovalum. On cultivated cocoa they were more at home at first, but after a while there were hardly any ants left and the mealybug numbers on all the host plants showed a decreasing trend during the experiment (Table 1). In the second experiment, six Theobroma-species and an unknown species of Herrania were tested. Seven replicate seedlings of each species were grown in plastic buckets and later infested with mealybugs and attending ants, Crem- atogaster africana (Mayr). The infestation took place by putting 20 females on each seedling every week during a period of 12 weeks. The ants were intro- duced in a large carton nest and they were found soon all over the seedlings. Population assessment of the mealybugs was started a fortnight after the last infestation and was repeated every fortnight up to 12 weeks. There were marked differences between the plant species tested and the cultivated cocoa was the best host plant among the species (Table 2). But again the ant numbers started to diminish and there was thus an overall decrea- se of mealybugs at the end of the experiment. It is evident that the success of a technique for mass-rearing of P. njalensis is dependent on providing favourable conditions for host plants, mealybugs Table 1. Food plant preference of P. njalensis. Mean number of adult females on seedlings at monthly intervals, months after infestation. Plant species and variety Months ■ 12345678 T. cacao L. Amazon x Cundeamor hybrid 63 44 52 31 23 15 9 11 Trinitario x Amazon hybrid 58 23 33 25 25 16 7 5 Inter-Amazon hybrid 54 32 34 27 19 15 15 9 Unselected Amelonado open pollinated 50 35 49 41 28 28 17 9 Amelonado x Amazon hybrid 30 19 35 19 17 14 8 4 T. grandiflora K. Schum 15 22 28 28 10 8 9 8 T. obovatum Bern 10 7 12 8 9 4 1 1 214 Table 2. Food plant preference of P. njalensis. Mean number of adult females on seedlings at two week intervals, weeks after infestation. ~, , . WeeksPlant speciesv 2 4 6 8 10 12 T. cacao L 117 180 189 208 199 130 T. speciosa Spreng 80 81 81 73 58 69 T. bicolor Humb. & Bonpl 39 44 45 44 36 21 T. microcarpa (Mart.) Sprague ... 31 33 40 49 40 67 T. grandiflora K. Schum 11 20 20 19 26 37 T. obovatum Bern 6 6 6 8 12 5 Herrania sp. A (indet) 1 6 5 4 4 8 and associated ants; in the absence of ants, mealybugs soon become smothered by moulds which develop on the honeydew. In these two tests such conditions could not be established for the attending ants. It is worth noting that several authors have tried to breed P. njalensis, or other mealybug species infesting cocoa, on different food plants with or without the attending ants, but always however, with discouraging results (Nicol 1950, Kirkpatrick 1950, Anon. 1951 a, 1953 a, Donald 1954 a, b, c, 1955, 1957, Decker 1956, 1957, Okusanya 1971). 2.3 Artificial feeding Mealybugs were sometimes fed on sucrose and other solutions in feeding tubes. These were glass tubes about 4.0 cm long and with an inner diameter of about 0.8 cm. A double parafilm sachet was fixed to one end of the tube and the other was closed with a foam polythene plug after introducing the mealybugs. Conical vacuum flasks were used sometimes instead of glass tubes. The sachet containing the liquid diet between the two sheets of parafilm was prepared as described by Mittler and Dadd (1964). 2.4 Feeding cones and cages In many experiments mealybugs were fed on cocoa seedlings by using the paper cone technique of Posnette and Strickland (1948). According to this technique, a conical piece of paper is fixed near the terminal bud of the seedling and the mealybugs are placed inside the cone. Because many mealybugs dis- appeared into the soil or otherwise escaped from the cones, this technique was replaced by confining the mealybugs into feeding cages which were fixed on the seedlings. A description of such feeding cages has been published by Markkula (1963). In this investigation the cages had an outer diameter of 2.5 cm, they were about 3.0 cm thick, and the diameter of the inner feeding space was 1.2 cm (Fig. 2). These cages had several advantages; the mealybugs could not escape, they were not exposed to external hazards, and they readily settled down to feed inside the cage (Fig. 3). 215 2.5 Physical conditions Handling of insects and other short term manipulations were usually carried out in air-conditioned laboratory rooms where the temperature varied between 20 and 28° C. depending on outside conditions. The relative humidity varied from 40 to 70 % in these rooms and the variation was indirectly proportional with the temperature. The experiments proper were done sometimes in an ordinary laboratory room with a considerably high temperature and humidity (Fig. 4). For most experiments, however, two growth chambers were used with standard conditions of 25—27° C., 55 —75 % of relative humidity, and about 2 000 foot candles of fluorescent light. Due to technical difficulties it was impossible to reproduce these standard conditions in some experiments. The deviations are shown in connection with those experiments. The standard conditions were arrived at after some experimentation described in the following. 2.5.1 Reactions of mealybugs to light Light reactions of mealybugs were investigated in air-conditioned laboratory at 25 27° C. and 40—60 % RH. A wooden box measuring 30 x 30 x 30 cm was used and the open top of the box was covered by a thick black cloth around Fig. 2. Two feeding cages for mealybugs attached around the stem of a virus infected seedling. 216 Fig. 3. Feeding mealybugs forming a dense mat around a stem of a cocoa seedling; a typical scene when the feeding cage is removed. Also note the presence of some larger adult females. the neck and head of the observer (ref. old-fashioned photographers). Inside the box on the floor there was a microscope lamp with a 12 V 5.2 W bulb in one of the corners of the box. The lamp gave a beam of light on the experimental area at about 15° angle from horizontal towards vertical. The experimental area was a horizontal piece of white paper on the floor of the box (Fig. 5). The area was divided into squares measuring about 1.6 cm 2 (0.25 inch2 ) each, and the set-up of squares was 9 X 17. Before mealybugs were introduced into the system, the temperature was measured in different places of the experimental area and at different distances from the light source. This was done with a copper-constantan thermocouple connected into a potentiometer (Foster Portable Potentiometer Model 3155- APX). This instrument could not detect any differences in the temperature at different places of the experimental area. Thus the amount of radiant heat must have been very small. Mealybug nymphs o—3 days old were introduced into the system one at a time and were liberated at the starting point which was in the middle of the beam of light at one end of the experimental area. The movements of the nymphs were recorded whenever they entered a new square of the area. This way of recording gave four possible directions of movement with equal chances, 217 namely, towards the light, away from the light, at right angles to the beam of light to the left, and to the right. Twenty nymphs of P. njalensis, P. citri and F. virgata were used in the experiment. The experiment was stopped each time when the nymph left the experimental area, and a new nymphs was intro- duced. The nymphs were obtained from reproduction cages kept in standard conditions but in dark for three days. All the three mealybug species behaved similarly. On the whole about 56 % of the movements were directed against the beam of light, 36 % of the movements were at right angles either to the left or right, and only 8 % of the movements were directed away from the light (Table 3). 2.5.2 Reactions of mealybugs to humidity Humidity reactions of mealybugs were studied in a growth chamber in controlled temperature conditions of 25—27° C. and with the lights off. Experi- ments were carried out in a linear gradient apparatus which was built according to Youdeowei (1967) but modified and smaller in size. The apparatus was a rectangular perspex box measuring 21.0 X 5.0 X 4.2 cm with a removable top and a false floor of 200 mesh wire gauze at 1.0 cm from the top. Below the false floor there were five dishes with graded aqueous solutions of sodium hydroxide (Madge 1961) to produce a humidity gradient in the experimental Fig. 4. Typical temperature and humidity conditions in an ordinary laboratory room. Recorded 19-26 June 1967. Fig. 5. Reactions of mealybugs to light. The experimental area with a track of movement as recorded («), starting point (S), and the beam of light indicated by the broken line. 218 Table 3. Movement of P. njalensis, F. virgata and P. citri in a beam of light. Frequency of movement observed towards given direction, mean of 20 nymphs, with expected frequencies in parentheses. Direction of movement At right At right Species and Towards angles to angles to Away from frequency the light the beam, the beam, the light right left P. njalensis Observed 159 61 49 30 Expected (74.75) (74.75) (74.75) (74.75) F. virgata Observed 158 40 43 11 Expected (63.00) (63.00) (63.00) (63.00) P. citri Observed 118 36 53 22 Expected (57.25) (57.25) (57.25) (57.25) chamber which measured 17.0 X 0.9 X 0.7 cm. The top was hermetically sealed with a rubber gasket, and small pieces of cobalt thiocyanate paper on the floor indicated different relative humidities (Solomon 1945). A central hole in the top, through which the insects were introduced, was sealed with a cover slip. Food was available above each hydroxide container as 2 cm long pieces of young cocoa shoots wrapped tightly in parafilm. The nymphs were obtained from reproduction cages which were kept in a growth chamber in standard conditions but in darkness for two days. In each test the nymphs were allowed to stay in the experimental chamber for 24 hours before the chamber was opened and the number of insects counted on different shoot pieces. The number of nymphs varied in each replicate test depending on the availability of nymphs. The nymphs reacted positively to high humidities and tried to avoid low humidities (Table 4). This happened with all the mealybug species tested and in every replicate test irrespective of different numbers of nymphs. 2.5.3 Reactions of mealybugs to temperature Temperature reactions of mealybugs were tested in a linear temperature gradient apparatus which was constructed according to Youdeowei (1968) but with some modifications. The gradient was composed of a wide U-shaped copper bar, 18 cm long, 2 cm wide and 1 cm deep. The bar was covered with a removable perspex top having nine equidistant holes for thermometers. A rubber gasket closed the system hermetically. Both ends of the bar were inside different water tanks, one with ice and water, and the other with an electric heater and a thermostat immersed in water. The whole system was insulated with porous polystyrene. The nymphs were obtained from cultures in reproduc- Table 4. Frequency distribution of P. njalensis, F. virgala, P. citri and P. longispinus in a humidity gradient apparatus at 25 —27 °C. with the preferred humidity in various places of the apparatus. Expected frequencies in parentheses. P. njalensis Relative humidity, % 10 30 50 70 90 Frequency, total of 6 replicates 142 394 406 721 1209 Expected (574.4) (574.4) (574.4) (574.4) (574.4) Relative humidity, % 95 20 95 20 95 Frequency, total of 4 replicates 1184 101 530 181 779 Expected (555.0) (555.0) (555.0) (555.0) (555.0) Relative humidity, % 20 95 20 95 20 Frequency, total of 4 replicates 18 458 71 685 72 Expected (260.8) (260.8) (260.8) (260.8) (260.8) F. virgata Relative humidity, % 90 10 90 10 90 Frequency, total of 6 replicates 21 7 24 5 19 Expected (15.2) (15.2) (15.2) (15.2) (15.2) Relative humidity, % 10 90 10 90 10 Frequency, total of 7 replicates 3 70 21 53 17 Expected (32.8) (32.8) (32.8) (32.8) (32,8) P. citri Relative humidity, % 10 30 50 70 90 Frequency, total of 8 replicates 11 31 33 24 153 Expected (50.4) (50.4) (50.4) (50.4) (50.4) Relative humidity, % 90 10 90 10 90 Frequency, total of 5 replicates 162 87 541 80 321 Expected (238.2) (238.2) (238.2) (238.2) (238.2) Relative humidity, % 10 90 10 90 10 Frequency, total of 5 replicates 56 107 26 132 82 Expected (80.6) (80.6) (80.6) (80.6) (80.6) P. longispinus Relative humidity, % 90 10 90 10 90 Frequency, totalof 2 replicates 86 15 38 6 38 Expected (36.6) (36.6) (36.6) (36.6) (36.6) Relative humidity, % 10 90 10 90 10 Frequency, total of 2 replicates 0 5 1 29 2 Expected (7.4) (7.4) (7.4) (7.4) (7.4) tion cages maintained in standard conditions in darkness for two days. These 0—2 day old nymphs were allowed to stay in the gradient for six hours. Ten 1 cm long pieces of parafilm covered cocoa shoots were put in between the thermometers for the nymphs to settle on and feed. The nymphs reacted more positively to temperatures of about 20—25° C. than to higher or lower temperatures (Table 5). It is possible, however, that the actual preferred temperature range is somewhat higher because the humidity 2 219 220 Table 5. Frequency distribution of P. njalensis nymphs in a linear temperature gradient apparatus. Temperature regime, C°. Frequency, total Expected of 12 replicates frequency 15.0-18.0 102 (123.9) 18.5-20.5 328 (123.9) 22.0-24.0 214 (123.9) 24.5-26.0 175 (123.9) 27.5-29.0 104 (123.9) 29.5-31.5 105 (123.9) 32.0-34.0 86 (123.9) 34.5-37.0 77 (123.9) 37.0-40,0 42 (123.9) 41.0-45.5 6 (123.9) was not controlled, and thus the nymphs may have responded more positively to the higher relative humidity of the lower temperatures. 2.5.4 Discussion and conclusions Information was required on reactions of virus transmitting mealybugs towards light, humidity and temperature to establish favourable feeding condi- tions in which virus transmission experiments could be carried out successfully. Posnette and Strickland (1948) report that ’all stages of P. njalensis are negatively phototrophic, and this response can be utilized for a bulk removal of these insects from cacao pods. The pods are placed in a Berlese-Tullgren funnel fitted with a 40 W electric bulb, and a collecting tube ... A large pro- portion of the nymphs and young adults migrate into the tube . . .’. In this case it is probably a seeming negative response to light because the insects may have responded to temperature rather than to light. Apart from this report there appear to be no further references dealing with reactions of virus transmitting cocoa mealybugs towards light, humidity or temperature. Detailed studies do exist, however, on some other mealybug species. Accord- ing to Salama (1970 a) the reaction of young females of Planococcus vitis Nied. towards light is a typical negative tropotaxis at 60 W illumination. In contrast, Parlatoria oleae (Colvee) is said to be a photopositive species, the reaction being reversed at high intensities of light (Hafez et al. 1967). The cocoa mealybugs tested in this study all reacted positively to light of low intensity. Although no specific experiments were carried out in other light conditions, it was observed from time to time that the intensity of 2 000 foot candles inside the growth chamber attracted nymphs of P. njalensis or F. virgata when the nymphs were kept on a black sheet of glass outside but near the illuminated observation window of the chamber. Moreover, inside the growth chamber in standard conditions, nymphs of P. njalensis settled down easily to feed in the feeding cages on cocoa seedlings. When the feeding cage was removed and the nymphs were exposed, most of them still remained settled for the next 24 hours. However, if these seedlings were taken outside the labor- 221 atory, then, in the presence of sunlight and heat, the exposed nymphs with- drew their stylets and moved down the stem to hide themselves in the soil in a matter of a few minutes. This reaction was slower on a half-cloudy day when the sun was not directly visible. Similar photic response towards direct sunshine has been reported by Salama (1970 b) for Mycetaspis personalus (Com- stock) in Egypt. It is possible, therefore, that several mealybug species respond positively to light of low intensity and negatively to high intensity light. Like some other insects, cocoa mealybugs display a definite response towards various relative humidities or different temperatures. No particular tempera- ture can be described as the optimum and we are rather dealing here with a wide zone of preference. Cocoa mealybugs appear to prefer high humidities when they are pie-conditioned at 60—75 % RH. According to Salama (1970 a) P. vitis has a very wide zone of thermopreferendum extending from 16 to 34° C. and, as far as can be judged, also this species prefers high humidities except when preconditioned at 95 % RH which resulted in the mealybugs showing indifference to humidity alternatives. 2.6 Temperature and reproduction Field collected females of P. njalensis were brought into laboratory where the adhering nymphs were carefully separated. Then these clean females were put into reproduction cages, 20 females into each cage, and the cages were kept in three different temperature regimes in the growth chambers for 44 hours. After this the cages were opened and the contents examined. There were marked differences between the three temperature regimes tested and the temperature of 25—27° C. appeared to be more favourable for reproduction by the fasting females than the higher temperatures. There were differences also in the rate of mortality of females and of nymphs in the three temperature regimes (Table 6). In another experiment, field collected females of P. njalensis were cleaned and put into reproduction cages at a rate of 20 per cage, and the cages were then kept in five different temperature regimes till the females were dead or moribund. The cages were opened daily and all the newly-borne nymphs were removed and counted. The cages were replicated four times, and there were four female populations collected from different places near the laboratory. Each population was tested simultaneously in two growth chambers in a control temperature of 25—27° C. and in one other temperature regime in order to check for population differences. Table 6. Reproduction of fasting females of P. njalensis at different temperatures during a 44 hour period. Nymphs Temperature Number of number number rate of survivors regime, C°. alive females alive dead reproduction per female 25-27 80 1484 9 18.718.5 31-33 80 805 69 10.910.1 36-38 39 39 356 4.90.5 222 Some reproduction occurred in all the temperature regimes but the highest yield of nymphs was obtained at temperatures between 25 and 33° C. (Table 7). There was at least some reproduction up to 13 days in the most favourable temperature conditions, however, most of the reproduction occurred during the first seven days (Fig. 6). Also, the reproducing females became moribund sooner in the high temperatures. The highest rate of reproduction during the first 24 hour period occurred at the highest temperature (Table 8). According to Strickland (1951) P. njalensis has a fecundity of 6—90 nymphs witha mean at 36. This is a strikingly low figure, though his experiments Table 7. Reproduction of fasting females of P. njalensis at different temperatures. Control temperature 25 27° C. with relative humidity at 55 —75 %. Temperature Relative Mean number of nymphs per regime, C°. humidity, % female ± S.E. x t 95 % 22-24 60-80 63.6 ± 20.2bl ) Control 96.9 ± 18.6° 28-30 55-75 114.7 ± 17.1<= Control 73.7 ± 52.9»bc 31-33 55-75 81.2 ± 28.8» Control 85.3 ± 38.9b 0 34-36 50-70 36.2 ± 9.7» Control 111.0 ± 20.0<= *) In this and some other tables, numbers followed by the same letter are not significantly Fig. 6. Reproduction of field collected P. njalensis females without food in different temperatures. 223 Table 8. Reproduction of fasting females of P. njalensis at different temperatures during a24 hour period. Temperature Relative Mean number of nymphs per regime, C°. humidity, % female ± S.E. x t 95 % 22-24 60-80 9.3 ± 3.3» 25-27 55-75 10.7 ± 2.6» 28-30 55-75 12.1 ± 3.7» 31-33 55-75 9.4 ± 3.2» 34-36 50-70 17.7 ± 4.6» were carried out in laboratory conditions, where the temperature and humidity were not recorded, and which may have been less favourable for reproduction. Besides, he used the term fecundity to indicate the maximum number of nymphs surviving after the first moult. In the present study the rate of reproduction was tested in different temperatures with field collected females which may have reproduced already while still in the field. Thus, the values obtained represent only a part of the total fecundity. However, these partial fecundity values average about 92 at control temperature of 25—27° C. and, therefore, the total fecundity in terms of number of live nymphs reproduced is probably well over a hundred in favourable conditions. This suggestion agrees with results obtained by some other authors on several mealybug species other than P. njalensi s (James 1937, Hafez and Salama 1970, Seuge et al. 1971). 2.7 Use of radioisotopes The use of radioisotopes was based on the early work of Cornwell (1957), who studied the uptake and translocation of 32P in seedlings and mature cocoa trees. In the present study, radioisotopes were used in many experiments and they were obtained from The Radiochemical Centre, Amersham, England. 32P was used as orthophosphate in a solution containing 1 mg/ml phosphorus buffer pH 7. 35S was in the form of carrier free sulphate in an aqueous solution, pH 6—B. Most often these commercial isotope products were used to label virus source plants to detect the subsequent feeding of mealybugs in various conditions. Virus infected cocoa seedlings were washed in tap water to clean the roots from soil and other debris and then the roots were immersed into complete culture solution in a 50 or 100 ml conical flask. Two layers of parafilm were used to seal the flask and fresh radioactive culture solution was added when necessary. The culture solution containing the radioisotope had a spe- cific activity of about 30 / F F(6) when I 0 == • then I = -Lt (7) when F = 0, then I = Ioe To facilitate calculation of constant L, T can be the half-life of L. Thus from (7) we obtain: 1 -Lt -l|t + T) (8) 206 = J»e which gives results: , m T dn 2) t (log 2) (9) L = or L= ■T T log e If log 2 == 0.30103, and log e = 0.43429 and, T = 12 hours, then L = 0.0578/hour. If Imax = F100 then, 100 = and, F = 5.78/hour. When these values are fitted into (4), then: (10) I== 100 [l (1 0,578 I 0) e-°-0578 t Fig. 29. Uptake of radioac- tivity by nymphs of P.njalensis feeding on 32P-active 10 % sucrose solution via a parafilra membrane. Black circles (») and open circles (o) indicate amount of food uptake in cplOOs by radioactive nymphs and all the nymphs, respect- ively. A; y = 30.4 + 5.025x; r = +o.B72***. B; y = -6.9 + 5.064x; r = +o.B9B***. 269 270 From formulae (7) and (10) it follows that when mealybugs have stopped feeding on virus source plant, then after 48 hours only 6.24 % of the infectivity remains. Also, when originally non-infective mealybugs have fed on virus source plant for 48 hours, then their infectivity is 93.8 % of the maximum in- fectivity. One specific experiment was carried out in a growth chamber at 25—27° C. and 50—75 % RH to obtain empirical data on validity of the formula (10). CSSV 1A isolate infected seedlings were used as virus source plants. Length of AAT varied from one to 120 hours and the age of the nymphs of P. njalensis was 0—3 days at the beginning of AAT. The length of lAT was 24 hours and three nymphs were put on each half bean test plant. Virus was not transmitted after a short AAT of I—6 hours but from 12 hours onwards the infection rate increased rapidly to reach a maximum at 48—72 hours (Table 35). A decrease in infection rate was noticed after an AAT of 96 and 120 hours. The experimental data are given with mathematical formulae in Fig. 30. In order to get a more general picture of the relationship between the linear food uptake and the exponential loss of infectivity, results were compiled from several experiments dealing with the length of the AAT and the infection rate (Tables 13, 14, 18—24 and 35). In each experiment, infection rate at 24 hours AAT was calculated close to 50 % by using the maximum likelihood estimator, and infection rates for other lengths of AAT were calculated accordingly. Thus the level of adjustment of infection rates remained the same within each ex- periment, but may have varied between different experiments. Finally, in- fection rate means were calculated for I—6, 7 12, 13—18, 19—24, 25 36, 37 60, 61—84, 85—108 and, 109—120 hours of AAT, and the infection rate means were plotted against time (length of AAT). Table 35. Effect of length of acquisition access time on the rate of CSSV 1A isolate transmission by nymphs of P. njalensis. Length of Infection Infection rate, AAT, hours rate % ± error1 ) P ’ 1 0/70 0 0 2 0/70 0 0 4 0/70 0 0 6 0/69 0 0 12 1/70 1 ± 4 0.5 18 25/70 36 ± 12 9 24 55/70 79 ± 10 40 30 45/70 64 ± 12 29 36 accidentally destroyed 42 45/70 64 ± 12 29 48 59/67 88 ± 9 51 72 58/63 92 ± 8 58 96 47/70 67 ± 12 34 120 47/68 69 ± 12 35 1) 95 % confidence limits for a proportion corrected for continuity. The infection rate means were in accordance with the mathematical ex- pectation up to 72 (61—84) hours AAT (Fig. 31). However, with 96 and 120 hours AAT there was a significant decrease in the infection rate, which is con- sidered to be due to the decreased amount of food uptake by prestarved mealy- bugs at later stages of prolonged feeding, or in other words, the disappearance of increased amount of food uptake caused by preliminary fasting (see 3.5). Fig. 30. The effect of length of acquisition access time (AAT) on the rate of CSSV 1A isolate transmission by nymphs of P. njalensis. Relationship (C, D) between linear uptake of food (B) and exponential loss of infectivity (A) with time. Black circles (») indicate infection rates ob- tained after different length of AAT with 95 % confidence limits for a proportion corrected for continuity. A; exponential loss of infectivity with time; I = 100e~°-°578 t; F = 0/h; I 0 = 100. B; linear uptake of food with time; F = 5.78t. C; increase of infectivity with time during AAT, a theoretical case with settling rate of 100 % at t 0; I = 100(1 - e -°- 0ä78t ); F = 5.78/h; I 0 =O. D; increase of infectivity with time during AAT, a case with observed settling rate of 50 % at t l2 ; I = 100 [1 - e —°-0578t t —12 >J; F = 5.78/h; 112I12 =O. 271 272 4. Discussion and conclusions 4.1 General Among the 15 families of Coccoidea (IMMS 1957), only the family Pseu- dococcidae, the mealybugs, have been shown to transmit plant viruses, although the general ecology and feeding behaviour is similar in all the families. P. cilri was the first mealybug species reported to be a vector of the tobacco mosaic virus (TMV) (Olitsky 1925), further confirmed by Newton (1953). Pseudo- coccus maritimus (Ehrhorn) was recorded as a vector of an unspecified mosaic disease (Elmer 1922, 1925). An unidentified mealybug species was reported to be a vector of a bean mosaic disease (Fajardo 1930). Hughes and Lister (1953) noted that lime dieback was transmitted by F. virgala. Mealybug wilt of pineapple was long considered to be caused by wilt in- ducing toxic secretions of Dysmicoccus brevipes (Cockerell) and D. neobrevipes Beardsley, however, in a recent reappraisal Carter (1963) has concluded that a transmissible latent factor is also involved which ’is presumed to be a latent virus’. Some of the above reports have not been confirmed, while in others there is not enough evidence to sustain virus entity, or no information is given on Fig. 31. Relationship between linear uptake of food (virus) by mealybugs and exponentia loss of infectivity (virus) in mealybugs. Black circles (») indicate infection rate means calculated from results of several experiments. A; exponential loss of infectivity with time; I = 80e~0.0578(t—9). p _ o/h; I 0 = 80. B; linear uptake of food with time; F = —41.62 + 4.62t. C; increase of infectivity with time during AAT, a case with Ig = 0; I = 80 [1 e 0.0578(t 9 )]; F = 4.62/h. D; suggested decrease of infection rate with long AAT due to starvation before AAT. 273 virus-vector relationships. Consequently, virus transmission by other homopte- rous insects, notably aphids, is the nearest parallel of mealybug transmission of cocoa viruses. It has not always been generally agreed that mealybugs are the sole vectors of cocoa viruses. During the search for vectors of cocoa viruses all other in- sects, particularly aphids and psyllids, which were at first claimed to be vectors of cocoa viruses (Cotterell 1943), failed to transmit virus in confirmatory tests (Box 1945, Posnette and Strickland 1948,Lister 1953) (see Appendix III). Recently, Martini (1961, 1962) repeatedly reported successful transmission of two isolates of CSSV by Aphis gossypii Glover in tests where 250 aphids were confined on 20 mm diameter areas on symptom bearing flush leaves for an AAT of 18 hours, and then transferred on half beans at a rate of 50 aphids per half bean; infection rates of 5 12 % were obtained. Longworth (1964 c, communication by letter 24 Jan. 1969) repeated these tests with the same tech- nique but no transmissions were recorded out of 376 tests. More recently, Eguagie (1970) used about 28 050 individual aphids with the same technique, with 18—36 hours AAT and 12—30 hours lAT. Virus symptoms developed in one out of 1 038 test plants and the single transmission in the experiment was attributed to mealybug contamination of test seedlings in the insectary. The conclussion thus seems evident that only mealybugs are the vectors of cocoa viruses. Relationships between mealybugs and cocoa viruses are analyzed in the following and, as much as possible, in the same sequence as different stages of virus transmission occur in nature. 4.2 Preliminary fasting/feeding According to Kirkpatrick (1950, 1953 a) fasting prior to AAT does not in itself increase the ability of P. citri to transmit CTV, but if the mealybugs are starved for some 6—24 hours, they tend to settle down to feed more readily on the source plant. Posnette and Robertson (1950), however, suggested that fasting before AAT increases temporarily the infection rate as well. They in- vestigated the effect of 12 hours of preliminary fasting on the settling and in- fection rate with P. njalensis and an AAT of 2—l2 hours on leaves of CSSV 1M isolate infected source seedlings. The settling and infection rates rose with the increasing length of the AAT and apparently more rapidly with starved than with unstarved mealybugs. Knowing that a high proportion of unstarved mealybugs would settle down to feed on cotyledons within 30 minutes, infected cotyledons were used as source plants in another similar experiment with AAT of 2 18 hours. The same but less noticeable difference was observed between the starved and the unstarved mealybugs. The infection rate was largely in- dependent of the settling rate when the two sets of mealybugs were compared, e.g. 48 % of the starved mealybugs settled at 10 hours AAT and gave an in- fection rate of 30 %, 47 % of the unstarved mealybugs settled at 12 hours AAT gave only an infection rate of 18 %. Thus in these experiments the infection rate increased through preliminary fasting but only part of the increase could be related to increased settling. 274 Dale (1954 d, 1955 a) tested the effect of preliminary fasting on the infection rate with CSSV 1A isolate infected source plants and single young adults of P. njalensis on test plants. The mealybugs were starved for o—4B hours and then given an AAT of 16 hours. There were no appreciable differences between the treatments; unstarved mealybugs infected 43 % of the test plants and mealybugs starved for various periods infected 25 —43 % of the test plants. Martini (1959 b, 1961) also used a constant length of AAT (2 hours) and various fasting periods of 15—120 minutes with P. njalensis and F. virgata. The starved and unstarved mealybugs gave similar infection rates from source plants that were infected with Balogun and Offa-Igbo isolates of CSSV. According to the present study, preliminary fasting increases the settling rate of P. njalensis during the first 6—B hours and probably up to 16 hours of feeding (see 3.5, Tables 11 14). Preliminary fasting also increased the rate of food uptake of the settled mealybugs. Starved mealybugs transmitted the virus more efficiently than the unstarved ones. All these findings are in accor- dance with results reported by Posnette and Robertson (1950), but not with those of Dale (1954 d, 1955 a) and Martini (1959 b, 1961). However, experimental details, e.g. number of test plants in each treatment, are lacking in these contradictory reports and, therefore, the right evaluation is impossible. It is concluded that preliminary fasting increases the settling rate and amount of food uptake of P. njalensis, and also that these two factors contribute to the increase of infectivity of the mealybugs; both increased settling and food up- take make it more probable that the vectors acquire virus. 4.3 Availability of virus in source plants Posnette and Strickland (1948) showed that 1A isolate of CSSV can be transmitted by P. njalensis while feeding on leaves, shoots or bark during the AAT. Young flush leaves appeared to be the best source of virus. Virus could be acquired by mealybugs also from pods on naturally infected trees. Dale (1958) found that the maturity of pods had little influence on the infection rate. The 1A isolate of CSSV was present in flowers of infected plants (Dale 1954 a, Anon. 1963). The same virus isolate could be transmitted from the testae of beans as well (Anon. 1950), but in numerous tests seed transmission was not obtained (Posnette 1947 a). Posnette and Robertson (1950) investigated the effect of symptom pattern and leaf age on the infection rate with 1M isolate of CSSV. Immature leaves with the red-veining symptoms provided the best source of virus for mealybugs, and slight differences in the maturity of leaves of the source plants could lead to considerable variation in the infection rate. Uniform infection rates were usually obtained when young seedlings were used as source plants but from older plants the results were often variable. The availability of virus appeared to decline as the plants became older. This might be expected with viruses which cause dissimilar acute and chronic phases like CSSV, but it was also noticed with CMLV 1C isolate, which does not (Posnette 1947 a). Few trans- missions with mealybugs were obtained from mature trees or from seedlings which had been infected with the 1C isolate for over a year, but once the virus 275 had been transferred into seedlings, subsequent transmissions were readily obtained from them. The availability of virus in different parts of source plant was investigated by Posnette and Robertson (1950). Separate parts of plants infected with 1M isolate of CSSV were colonized by mealybugs and comparable tests were done. Although normal infection rates were obtained from 22 feeding sites, no transmission was obtained from four sites. Three of these were petioles and the fourth was a stem. The probability that the lack of transmission by these colonies was due to chance was 0.005 for each of them and this suggests that systemic infection was incomplete. The relative transmissibility of mild and virulent isolate of CSSV 1A isolate by P. njalensis was investigated by Posnette and Todd (1955). In contrast to the virulent isolate, which was transmitted usually to over 30 % of the test plants, the mild isolate was transmitted at widely varying rates according to the site at which the virus was acquired, but never was the mild isolate more readily transmit+ed than the virulent. From newly infected seedlings, the two isolates were transmitted almost equally at first, but the mild isolate became much less readily transmitted as the symptoms became less conspicuous in the second and later flushes. Substantiated by detailed studies Thresh (1958 c) indicated that CSSV is more extensively distributed in available form in infected trees with symptoms than in trees without them. When mealybugs were allowed to feed on samples collected from obviously infected trees, they later infected 170 of the 381 test beans (44.6 %) which were colonized, whereas mealybugs which had fed on the samples from latently infected trees, infected only 76 out of 332 (22.9 %) beans infested. In an experiment using 1A isolate of CSSV and two year old Iquitos type seedlings with symptoms of varying intensity as virus source plants and P. njalensis as the vector, the following infection rates were obtained: severe symptoms 38/215 (17.7 %), intermediate symptoms 30/282 (10.6 %), regular mild symptoms 21/145 (14.5 %), very mild sporadic symptoms 1/248 (0.4 %) (Dale 1958). Igwegbe (1966 a) compared infection rates using CSSV infected seedlings in acute and chronic phases of infection and P. njalensis and F. virgata as vectors. Regardless of the feeding site, considerably higher infection rates were obtained from source seedlings in the acute phase of infection with both vectors. Owusu (1972) studied the availability of CSSV 1A isolate in recently in- fected Amelonado cocoa seedlings and mature trees with P. njalensis. It was established that soon after inoculation and throughout the greater part of the latent period, both seedlings and mature trees were potential sources of virus to mealybugs. Two mature trees yielded virus without ever producing disease symptoms, while two other plants which produced symptoms, never yielded virus. In the present study, higher infection rates were obtained with P. njalensis while feeding on stems than on leaves of young infected cocoa seedlings (Table 15). However, this may reflect feeding behaviour of the species rather than actual differences of virus availability as discussed in more detail under 4.8. 276 There was no relationship between the age of the source seedlings (29 —54 days) and the infection rate (Table 16). Significant random variation was observed between young source seedlings of the same age (Table 17). The chance of mealybugs becoming infective depends on the availability of virus in the tissues on which the mealybugs are feeding. It is concluded that in general mealybugs can pick up virus from any part of plant except the seed in untouched pod. There is evidence that virus is not always fully systemic in the host. Recently infected plants in the acute phase of infection and with severe symptoms are good sources of virus while availability decreases with age of infection when the infection turns into the chronic phase. Availability of virus is not always indicated by the presence of visible symptoms. 4.3.1 Availability of virus in different cocoa types It is generally known that tolerance to virus exists among certain types of cocoa. A tolerant plant produces mild and often transient symptoms of infection and this would render detection of infection with CSSV very difficult.lmplicit in their use, therefore, is that the control of virus by eradication methods would be almost impossible. It would be of value if virus was less readily available in tolerant cocoa types, as this would tend to restrict spread of infection in a tolerant population and from it to a susceptible population. Dale (1958) used virus tolerant Iquitos-type cocoa seedlings infected with 1C isolate of CMLV and similarly infected Amelonado-type cuttings as controls to investigate the availability of virus for P. njalensis. The infection rate from tolerant Iquitos was 44 % and from susceptible Amelonado 76 %. Brunt (personal communication reported by Blencowe 1962) found that when batches of 25 nymphs were transferred to test plants after feeding on twigs of infected Amelonado trees, relatively intolerant Amazon trees, and tolerant Amazon trees, the infection rates were 43/89 (48.3 %), 28/95 (29.4 %) and 16/93 (17.2 %), respectively. Availability of Egbeda isolate of CSSV was examined by Longworth (1964 b) in seven populations of Amazon hybrids with two known to be tolerant and with Amelonado-type cocoa as control. P. njalensis crawlers were used as vectors and beans as test plants. The infection rates varied from 0 to 70 %, and the results were considered inconclusive. In a repeated test the availability of virus was highest in the Amelonado control and one of the hybrids not tole- rant of infection. Otherwise the results were variable; even high availability of virus in some replicates of tolerant cocoa was observed and sometimes virus was less available in sensitive cocoa. No transmissions whatever were recorded from three trees, although the trees had shown symptoms of virus and quite a large number of mealybugs had been used. In another test the availability of Offa-Igbo and Egbeda isolates of CSSV was followed in three cocoa hybrids and with Amelonado-type cocoa as control. No general pattern emerged, ex- cept that in many plants availability was high only during two months after the date of infection and fluctuated in subsequent tests, often widely. Tinsley (personal communication reported by Longworth 1964 b) followed the availability of two isolates of CSSV in seven different cocoa types in seedlings 277 derived from open pollinations which were infected as beans. Availability in the first nine months after germination was high in all types (86 —99 %), pre- sumably as the virus was completely systemic and the seedlings were in the acute phase of infection. At 12 months, when the seedlings were in the chronic phase of infection, availability was greatly reduced (I—4o %). Igwegbe (1966 b) studied the availability of Egbeda isolate in two suscep- tible and four tolerant or suspected tolerant cocoa types with P. njalensis. There were no consistent differences between the cocoa types in virus availability, however, it was noted that two types of cocoa in the chronic phase of infection after 22 weeks gave very low infection rates, 0 and 5 %, whereas the other four types not yet in the chronic phase gave infection rates of 15—79 %. Five tolerant and sensitive cocoa trees were used by Legg and Bonney (1968 b) to investigate the availability of 1A isolate of CSSV for P. njalensis. The tolerant Amazon-type trees had good canopies and showed only mild leaf symptoms, while the sensitive Amelonado trees had severe leaf symptoms and extensive dieback. The infection rate from tolerant trees ranged from 4 44 % (mean 17.2 %) and from sensitive trees from 472 % (mean 24.9 %) however, the mean differences for all tests were not significant. Owusu (1969, 1970, 1971 b, 1973) reported a series of trials to investigate the availability of virulent 1A isolate of CSSV to the mealybug vector P. njalensis in virus tolerant and virus sensitive cocoa. Acquisition feeds were given directly on young plants in the greenhouse and on mature trees in the field without severing the branch. In the first trial with mature cocoa of 7—17 years there were five tolerant Amazon trees with five Amelonado trees as controls. Avai- lability of virus was investigated five times in these trees at about monthly intervals over a period of six months. The infection rate from the sensitive trees was 1040/1963 (53 %) and from tolerant trees 603/1967 (31 %). In the second similar trial, but with eight replicate trees of each sensitive and tolerant type, the infection rates from sensitive and tolerant trees were 996/2223 (45 %) and 627/2297 (27 %), respectively. In another trial with six replicate young seedlings of each sensitive and tolerant type there was no difference between the infection rates; 227/774 (36 %) and 253/731 (35 %), respectively. In a repeated trial with cocoa about two years old an infection rate of 744/1985 (37 %) was obtained from sensitive cocoa and 540/2137 (25 %) from tolerant cocoa. It can be concluded that virus is more easily available or more readily trans- mitted by mealybugs from sensitive than tolerant cocoa, however, the differ- ence in availability is relatively small. At most the difference appears to be the same as that between infection rates obtained by using one or two mealybugs per test plant. 4.4 Acquisition feed 4.4.1 General conditions during acquisition feed Despite all efforts to standardize conditions during AAT, or during the ex- periments, there were sometimes significant differences between the infection rates of individual virus source plants in the same treatment (e.g. Table 17). 278 From time to time such differences occurred even though standard conditions of temperature, humidity and light were maintained during the experiment, the source plants were of the same age and displayed similar acuteness of sym- ptoms, and radioisotopes were used to verify the uniform feeding of mealybugs. Although this ’innate’ difference between replicates could be overcome by increasing the number of source and test plants, it still remained an unnecessary nuisance in the experimental work of the present study. Kenten (in the press) has recently investigated conditions during AAT and shown that source plants grown in sand and watered with nutrient solution may differ less than plants grown in soil with tap water. The range of infection rate of Amelonado test beans from 11 randomly selected source plants grown in soil was 24.5—84.8 % and from those in sand 40.3 72.7 %, with coefficients of variation of 23.5 % and 13.1 %, respectively. The difference between the two groups was significant. Occasionally, owing to shortages, source plants have to be used twice or more often for an AAT. No significant difference was found in the infectivity of mealybugs fed on source plants which were used once, twice or thrice for an acquisition feed; the infection rates were 58.3, 57.6 and 63.0 %, respectively. During acquisition feed the nymphs are kept in feeding cages (see 2.4) close to the stem of infected seedlings and are buffered to some extent against the effects of low humidity. Kenten (in the press) found no difference in the infectivity of nymphs fed at an external humidity of 35 % or 90 %, the infection rates were 74 % and 65 %, respectively. As described under 3.7.1, mealybugs are removed from source plants by tapping the plants repeatedly, however, the mealybugs that first fall off the plants may not have fed at all and batches of nymphs obtained through later tappings contain a higher proportion of mealybugs that have settled down to feed during the AAT. This was indicated by a higher proportion of radioactive mealybugs in later tappings. Kenten (in the press) has now obtainedevidence that there is also a difference in infectivity between batches of mealybugs obtained through repeated tapping. The mean percent transmission to Amelonado test beans, after angular trans- formation, was 38.0 % with the first batch, with the second 44.9 % and the third 49.7 % (L.S.D. 5.2 at P = 0.05). Such variation was avoided by tapping the nymphs off the plants repeatedly into a pool of nymphs on a large clock glass. Using these clock glasses and pools of nymphs in closely similar conditions in four tests, standard errors of the mean percentage transmission were satis- factorily low for all the five persons manipulating the mealybugs, yet the number of transmissions in one test was less than half of those obtained in the other three but the source of this variation could not be identified. 4.4.2 Length of acquisition feed With different cocoa viruses the infection rate generally rises with the in- creasing length of AAT. Posnette and Strickland (1948) used 1A isolate of CSSV and P. njalensis to test the effect of an AAT of 4—48 hours on the rate of virus transmission. The infection rate increased with the length of the AAT and the highest rate of transmission was obtained with an AAT of 48 hours. 279 Posnette and Robertson (1950) tested starved and unstarved populations of P. njalensis and 1M isolate of CSSV in two experiments. The length of the AAT was 2 12 hours in the first and 2 18 hours in the second experiment, and an overall increase of infection rates was recorded with increasing length of the AAT. Dale (1954 c, d) used P. njalensis starved for 18 hours before the AAT and CSSV isolate 1A in one experiment where the length of the AAT was 2 64 hours, and in another with an AAT of 8—32 hours. The infection rate rose with an AAT up to 16—32 hours but apparently decreased with long AAT of 64 hours. In a subsequent experiment with starved individuals of the same species the length of the AAT was I—6 days at one day intervals and the in- fection rates were 80—90 % in all the treatments (Dale 1954 d). Dale (1958) also reported infection rates of 43—55 % with the length of AAT at 1, 2 and 3 days. After a short AAT of one hour P. njalensis did not transmit the 1A isolate of CSSV (Dale 1958), and only few transmissions occurred after an AAT of 90 minutes (Dale 1957). Posnette and Strickland (1948) experimented with 1A isolate of CSSV and F. virgala. The length of the AAT was 4 and 48 hours and the infection rates were 1/29 and 6/27, respectively. With virus CTV and P. citri as the vector, the minimum length of the AAT required for transmission was 33 minutes. An AAT longer than one hour did not substantially increase the infection rate (Kirkpatrick 1950). In the present study the effect of the length of the acquisition feed was studied in several experiments (see Tables 13 14, 18—24, 35). A steady impro- vement in terms of higher infection rates was experienced during the five yeax experimental work of this study. This was largely due to increased knowledge of favourable transmission conditions and material improvements, e.g. optimum temperature and humidity, arrival of two growth chambers, more skilled hand- ling of mealybugs and other improvements related to the experimental techni- que. A direct comparison of results is therefore difficult. However, through the use of maximum likelihood estimator such a comparison can be made because infecton rates in different experiments are brought to the same level without altering the trends of results. In general, the results of the present study are in agreement with those reported by other authors, and there is enough accumulated evidence to conclude that with CSSV 1A isolate and P. njalensis as the vector, the infectivity of mealybugs increases with the length of the AAT and maximum infectivity is reached with an AAT of 48 —72 hours. The settling rate of mealybugs causes a delay in the actual commencement of feeding and thus the optimum length of the acquisition feed proper must be shorter. With a longer AAT of 96 hours or more the infectivity decreases with mealybugs that have gone through a preliminary fasting period long enough to cause considerable increase of settling and feeding in the beginning of the AAT (see Fig. 31, D). The contradictory evidence reported by Dale (1954 d), namely that equal and high infection rates were obtained with an AAT of I—61 —6 days, is explained by the excessively high level of the infection rate which obscured differences between treatments. 280 Although there is less information on other West African virus isolates or other mealybug species, no controversial evidence has been obtained in this paper or elsewhere to the general statement given for CSSV 1A isolate and P. njalensis. The single report of Kirkpatrick (1950) indicates that the effect of the length of the AAT on virus transmission by mealybugs may differ between the West African viruses and the CTV. 4.4.3 Acquisition of virus from liquid leaf extract Although no conclusive evidence was obtained in the present study to the effect that virus can be acquired by mealybugs while feeding on liquid leaf extracts via a parafilm membrane (see 3.7.5), some space is devoted to discussing this approach and the conditions involved. It is a well established fact that CSSV and CMLV can be transmitted mechanically (Brunt and Kenten 1960, 1962). As an example in this process about 0.5 kg of fresh virus infected leaves are required to produce 1 ml of infective virus preparation. With suitable abrasives this 1 ml is applied on 20 cocoa test beans and usually about half of the beans become infected. Thus, a relatively large volume of 0.1 ml of virus preparation is needed to infect one bean. Young nymphs of P. njalensis, o—2 days old, are minute creatures weighing only 0.0094 mg each (mean of 266 nymphs). Such a young nymph is able to take up food at a rate of about 0.005 /tl/day while feeding on 10 % sucrose solution (see Tables 11, 12). On cocoa seedlings the amount of food uptake may be higher but not so much as to cause relevant errors in this calculation. Assuming that a normal infection rate of 20 % is obtained with an AAT of 24 hours and single nymphs are used on test beans during lAT, it follows that mealybug nymphs require about 0.025 [A of infected plant sap to infect one bean, and the actual volume injected into the bean must be only a fraction of this. It is evident, therefore, that by volume the 'mealybug technique’ of virus transmission is at least 4 000 times more efficient than the technique of mechanical transmission. The reasons for such a difference should be investigated. It is possible and plausible that either, (1) the mealybug inoculation of virus with mouthparts into the beans is more efficient than the mechanical inoculation of brushing with abrasives, or (2) the inoculum injected by mealybugs and the liquid leaf extract used in mechanical transmission are not comparable as regards the infective virus titer. That mealybugs did not transmit the virus into beans with ease after feeding on the liquid leaf extract indicates that the low titer of infective virus in the extract may be one of the possible reason, however, it is not excluded that the chemicals in the extract may have had an adverse effect on the mealy- bugs and their transmission efficiency. Okusanya (1969) reported recently that P. njalensis fed via parafilm on liquid leaf extracts failed to transmit CSSV. None of the 230 test beans became infected. Okusanya (1969, 1970) also reported two tests on successful trans- mission of cocoa virus from liquid leaf extracts by P. njalensis. Three different liquid leaf extracts were prepared from CSSV Ife, Ikire or Offa-Igbo isolate infected cocoa leaves and the extracts were mixed with 4 % agar at 40° C. The mealybugs were given an AAT of 16—24 hours on the agar and an IAT 281 of 18—24 hours on beans with 15—20 crawlers on each bean. In the first test infected plants were obtained from two isolates, Ikire and Offa-Igbo, which gave infection rates of 76/112 (68 %) and 36/74 (49 %), respectively. In the second test Ikire and Offa-Igbo isolates were used again and the infection rates obtained were 363/621 (58 %) and 390/614 (64 %), respectively. However, none of the infected test plants in these tests showed the typical CSSV symptoms and threrefore Okusanya (1969, 1970) suggested that a new virus, the cocoa chlorotic mottle virus, was involved and that the new virus was an isolate of CSSV. It appears to be too early to decide upon the right status of these new research findings before further evidence from confirmatory tests is avai- lable. 4.5 Persistence of virus in mealybugs after acquisition feed 4.5.1 Persistence of virus in fasting/feeding mealybugs Posnette and Robertson (1950) used a mixture of all stages of ’fully fed’ P. njalensis to test the persistence of CSSV 1M isolate after AAT by starving mealybugs. The infectivity declined gradually up to 20 hours of fasting and then more rapidly, and the mealybugs infected one test plant after 34 hours of fasting. With the same virus isolate and P. cilri as the vector, there was no decline in the infection rate up to 18 hours of fasting, but after this infectivity declined and the virus did not persist longer than 36 hours in the fasting mealy- bugs. Lister (1953) found that the 1A isolate of CSSV persisted in starving adults of P. njalensis up to 49 hours and in first instar nymphs up to 23 hours. In Trinidad the CTV persisted in P. cilri up to 23 hours and in D. brevipes up to eight hours after leaving the infected plant (Kirkpatrick 1950). Martini (1961) reported that mealybugs bred on infected plants retained their infectivity unchanged up to 17 hours if starved on glass. After 24 hours of fasting, infectivity declined to about 50 % and was lost entirely after 30—40 hours. However, the infectivity of insects was lost after 17 hours on glass when a short AAT of four hours was given. If postacquisition fasting was done on filter paper, the infectivity was lost more rapidly than on glass. Carter (1961) kept P. njalensis after an AAT of 48 hours on 3 % agar for various periods before lAT. Infectivity of the mealybugs persisted up to 24 hours, but it is not known whether the mealybugs actually fed or voluntarily starved on the agar. Longworth and Entwistle (1965) allowed infective mealybugs to probe into agar, subsequently as the stylets were withdrawn they were passed through tannic acid solution; no transmissions were obtained. In another experiment infective young adults of F. virgata were confined in special agar cells and the stylets were observed as they probed the agar. After probing continuously in agar for up to three hours the mealybugs were still infective. Posnette and Strickland (1948), transferred infective P. njalensis on test plants for an IAT of 3—192 hours, after which the insects were transferred 282 to a second series of test plants for an IAT of seven days. None of the plants in the second series became infected. Posnette and Robertson (1950), using CSSV 1M isolate and P. njalensis, transferred single infective mealybugs to fresh test plants at intervals of 30 minutes. On only three occasions did an insect infect two plants in a series. One of the insects infected the first test plant, failed to infect the second, but again infected the third test plant in the series. The other two insects infected the first and the second plants in series. Also, in the present study, the infectivity of mealybugs declined gradually after AAT (3.8). This decline was best described as exponential with time. In terms of a maximum the virus did not persist longer than up to 96 hours. Persistence of CMLV in P. njalensis and CSSV in F. virgata were closely similar to persistence of CSSV in P. njalensis. These results agree largely with those reported by other authors, however, infectivity did not remain unchanged up to 17—lB hours of postacquisition fasting as reported by Posnette and Ro- bertson (1950) and Martini (1961), but rather started to decline earlier. Evidence was obtained in the present study that feeding after AAT increases virus persistence rather than decreases as reported by Martini (1961). It is unfortunate that details of experimental procedure are not described in that report. It is certain that the infectivity of mealybugs declines gradually with time after leaving the virus infected plant, and after about three days there is little infective virus left in the mealybugs. It is doubtful whether postacquisition feeding on healthy plants has any effect different from postacquisition fasting on virus persistence as such, yet postacquisition feeding may cause a delay in the settling to feed on test plants and thus reduce the infection rate. 4.5.2 Persistence of virus through postacquisition moulting The critical test of whether a virus is transmitted after a moult is considered to give conclusive evidence on the mechanism of virus transmission. If virus is transmitted after a moult the mechanism is circulative, if not, the virus is stylet-borne, for during a moult the stylets with any virus contamination are cast. Martini (1959 a) used Nigerian isolates of CSSV and nymphs of F. virgata and P. njalensis taken from infectedplants during moulting, but no transmissions were obtained with these insects. Longworth and Entwistle (1965) placed single infective nymphs of F. virgata in watch glasses and examined them for absence of exuvial skins before they were left for 12 hours to moult. After this period the bugs were examined again and moulted nymphs were transferred singly to cocoa half beans for an lAT. Double the number of unmoulted nymphs were used on beans as controls. Altogether some 100 moulted nymphs were obtained, but none was infective. However, transmission rates with the unmoulted control in- sects were so low as to make comparison impossible. Successful transmission of CSSV 1A isolate by moulted nymphs of P. njalen- sis was reported by Roivainen (1971). Not more than one day old nymphs 283ti were given an AAT of 120 hours on infected seedlings and after this they were transferred into moulting cages, one nymph per cage, for 20 hours. Moulting was indicated by presence of exuvial skins in the cages, and the moulted and unmoulted nymphs were put on separate half beans at a rate of I—s per bean. After an IAT of 24 hours the nymphs were killed, the beans were planted and the germinated seedlings were examined for virus symptoms. In seven tests the moulted insects gave a total infection rate of 35/259 (14 %) and the unmoult- ed control insects 45/254 (18 %), the difference was not significant. In the present study a test is reported on the transmission of CMLV 1C isolate by moulted nymphs of P. njalensis (see 3.8.7), however, only one test plant out of 65 was infected by moulted nymphs and six out of 65 by the un- moulted ones. In accordance with discussion by Longworth and Entwistle (1965), and Roivainen (1971), experimentation with moulted mealybugs was limited by several factors in the present study, e.g. partial loss of infectivity during the 20 hour period in moulting cages, generally low and verjf erratic rate of moulting, and excessive handling of mealybugs. However, with a large material consistent results should be obtainable. It is not known whether settling and feeding behaviour of moulted and unmoulted mealybugs on test beans differs to such an extent as to make direct comparison of infection rates unjustified. It is suggested that CSSV 1A isolate is transmitted after a moult and thus the virus is circulative in the vector P. njalensis. More data on the transmission mechanism of cocoa viruses in their mealybug vectors is highly desirable. 4.6 Inoculation feed Virus infection is manifested in symptoms of virus in the test plants after inoculation feed. With 1A isolate of CSSV, accumulated experience has shown that Amelonado-type cocoa seedlings, infected as beans, show symptoms of virus on the first or the second leaf flushes, or 17 —25 days and 40—5O days after planting of the beans, respectively. Symptomless seedlings can be con- sidered healthy (Posnette and Strickland 1948). Such ’healthy’ seedlings can be used to calculate the infection rate, but they cannot be considered as healthy for any critical experiment because occasionally symptoms develop during later flushes (e.g. 3.9.1). It is generally accepted that symptom devel- opment in this respect is largely the same in seedlings infected by other cocoa viruses. Although the length of the lAT or the number of mealybugs on test beans during lAT may be considered as obvious factors influencing the final infection rate, there can be other factors which are less obvious, e.g. Kenten (in the press) found that the moisture status of cocoa test beans influences the infection rate, probably because of mealybug preferences towards moisture. No difference was detected between infection rates obtained with beans from over-ripe or under- ripe pods. In the present study the location of mealybugs on the test beans at the end of lAT did not affect the infection rate. 284 4.6.1 Length of inoculation feed Posnette and Strickland (1948) used P. njalensis and isolate 1A of CSSV while testing the effect of length of the lAT on the infection rate. The length of the lAT was 3—192 hours. No clear relationship was found between the in- fection rate and the length of the lAT; by and large the infected test plants were randomly distributed with some infected test plants almost in all the treatments. The length of the lAT and its effect on the infection rate with 1M isolate of CSSV and P. njalensis was investigated more in detail by Posnette and Robertson (1950). In the first experiment the length of the lAT was 1— 12 hours and the infection rate varied only between 32 and 40 %. In the second experiment the length of the lAT was reduced to 10 —6O minutes with 10 mi- nute intervals. First infections occurred at 20 minutes and the infection rate increased with the length of the lAT up to 50 minutes. Dale (1954 a, b, 1955 a) carried out similar experiments with the 1A isolate of CSSV and P. njalensis. The vectors could often infect a test plant after an lAT of 15 minutes, and from this the infection rate rose rapidly with increasing length of the lAT up to one hour and more slowly up to four hours. When the mealybugs settled rapidly, maximum infection rate was obtained with an lAT of two hours. Martini (1959 a) reported that P. njalensis transmits most frequently during the 45 —75 minutes of lAT, and only few transmissions occur during the first 30 minutes or after 150 minutes on the test plants. F. virgata transmitted the CSSV 1A isolate during an lAT of six hours (Posnette and Strickland 1948). In experiments by Kirkpatrick (1950) with the CTV, an lAT of 37 hours was as effective as one of 18—24 hours. Shorter periods were not investigated in detail but transmissions were obtained after an lAT of 90 or 100 minutes. The available data indicates that the inoculation access threshold is about 15 minutes. Probability of infection rises rapidly during the first hours of the lAT up to about three hours and there is usually no further increase after longer inoculation feeds. In the present study the settling rate of mealybugs on cocoa seedlings or 10 % sucrose solution reached a maximum in 8— 24 hours (see Tables 11, 12, 14, 24, 34), and although mealybugs may settle down to feed on half beans faster (Posnette and Robertson 1950) it is possible, in view of the short lAT required for maximum transmission, that mealybugs transmit virus into cocoa beans also during short probes preceding actual feeding on the beans. 4.6.2 Number of mealybugs on lest plants during inoculation feed Using IM isolate of CSSV and P. njalensis, Posnette and Robertson (1950) found that the infection rate rose from 18 % with one insect to 53 % with five and to 77 % with ten insects per test plant. When the number of insects was increased from one in steps of five up to 30 per test plant, the infection rate rose from 10 to over 90 % with the isolates 1A and 1M of CSSV. 285 Dale (1955 b) used 1A isolate of CSSV and P. njalensis in two experiments- With 1,2, 4,6, 8 and 10 mealybugs per test plant, the infection rate rose from 18 to 92 % in the first experiment and from 44 to 97 % in the second, while the increase of infection rate was relatively faster with low insect numbers. Adegbola (1965), while testing the relative efficiency of P. njalensis and F. virgata, used Egbeda isolate of CSSV and 10, 20 and 30 crawlers per test plant. The infection rates for the two species were 12 and 6, 19 and 26, 30 and 28 %, respectively. In experiments with F. virgala and 1A isolate of CSSV, Posnette and Strickland (1948) obtained an infection rate of 80 % with five insects on test beans. According to Kirkpatrick (1950) the virus CTV was transmitted by P. ciiri at a rate of 14—32 % with I—B mealybugs per test plant. It is concluded that the infection rate generally increases with the number of insects used during the lAT, as demonstrated by several authors and including the present report (see Tables 25, 26, 28 30), although the increase may be obscured by variation in and between tests as experienced in the present study, and e.g., also by Kenten and Legg (1970). The infection rate increases with rising insect numbers according to the laws of probability as demonstrated by Posnette and Robertson (1950) and confirmed in the present study (sec 3.9.2, Fig. 26). Therefore, there is no mass action effect of virus involved, or in other words, individual mealybugs do not inject subminimal doses of virus which can accumulate to produce infection when groups of these insects are used. Selection of the right number of insects for inoculation of test plants is an important factor because it influences the sensitivity and statistical signif- icance of the results. This was demonstrated under 3.9.3. Among the numerous papers, including the present one, where the number of mealybugs on test plants and the infection rate is reported, the probability of transmission with single mealybugs shows wide variation. However, with the accumulated knowledge on optimum conditions for transmission a probabil- ity of about 0.2 —0.3 can be maintained, and with more elaboration of ex- perimental technique a probability of 0.5 or more can be achieved (Dale 1958) (see Table 35). 4.6.3 Inoculation feed with virus resistantjtoleranl cocoa Posnette and Todd (1951) were the first to show that some Upper Amazon cocoa types are more resistant to infection and more tolerant of infection with CSSV than the Amelonado-type. This observation is now generally accepted. Mealybugs have been used as vectors of virus to show that some cocoa types are more difficult to infect than others. A good example of such investigations is given by Blencowe and Attafuah (1959) as follows. When 50 CMLV 1C isolate infected mealybugs were fed on each of 250 fan cuttings taken from Amazon-type cocoa the infection rate was 119/250. At the same time, and with similar technique Amelonado cocoa was infected at a rate of 19/20. Similarly, when three CSSV 1A infective mealybugs were fed on each of 380 Amelonado beans the infection rate was 63 %, while an equal batch of Amazon 286 beans was infected at a rate of only 43 %. Also, with batches of Amelonado and Amazon beans, five infective mealybugs per bean were required to infect 50 % of the Amelonado beans, whereas nine mealybugs were needed to attain the same level of infection with the Amazon beans. In the present study, possibilities in the use of P. njalensis for resistance and tolerance testing were demonstrated under 3.9.3. However, since the original discovery of mechanical transmission of CSSV by Brunt and Kenten (1960), suitable techniques have been developed to test resistance and tolerance with manual transmission of virus. According to Kenten and Legg (1970) manual inoculation technique not only improves the sensitivity of the estimation of resistance but also considerably simplifies such studies, because the mealybug method is so difficult to standardize and is so unreliable that it is impracticable to use mealybugs for largescale tolerance and resistance studies. It is obvious that the mealybug technique of virus transmission suffers from many known and unknown factors that may be difficult to standardize, and all these factors contribute to variation experienced between replicates and tests, however, it is as yet not known why the manual method and the mealybug method do not seem to measure resistance and tolerance in the same scale. For example, Kenten and Legg (1970) obtained almost similar infection rates of 13.3 and 12.6 % with the manual and the mealybug method, respecti- vely, while testing a resistant and tolerant Amazon-type. In the same exper- iment a cocoa type of intermediate resistance and tolerance was infected at a rate of 50.0 and 19.5 % and the susceptible Amelonado gave infection rates of 95.1 and 32.2 % with the two inoculation methods, respectively. It is appar- ent that the manual method overestimates the resistance/tolerance value of the Amazon-type in comparison with that of the Amelonado, and although the manual method will probably yield highly resistant and tolerant cocoa types (Legg and Kenten 1970), the final evaluation must be done with mealy- bugs which transmit the virus in the field. 4.7 Effect of vector age on virus transmission Posnette and Strickland (1948) suggested that crawlers ( = first and second instar nymphs) of P. njalensis may be more efficient vectors of CSSV 1A isolate than older nymphs or adults. Ten to 14 crawlers infected 31 %, 5 8 older nymphs 12 % and 4—5 adults 13 % of the test plants. Dale (1955 a, b) used CSSV 1A isolate and three insects (P. njalensis) per test plant in similar experiments. The infection rate was 68 % with crawlers, 74 % with older nymphs and 75 % with young adults. With single insects on test beans, adults infected 57 % and nymphs 54 % of the test plants (Dale 1958). If unskilled assistants carried out the transfer of insects and two adults were compared with six nymphs, the infection rates were 75 and 95 %, respec- tively. Similarly, five adults infected 26 % of the test plants and 15 nymphs 61 %. These tests indicated that unskilled handling did not favour the more robust adults (Dale 1957). 287 Posnette and Robertson (1950) compared adults and nymphs of P. cilri by using five insects per test plant. The adults infected 47 % of the test plants and the nymphs 29 %. The difference was significant at 5 % level. With the virus CTV and P. cilri as the vector, Kirkpatrick (1950) obtained transmission rates of 33 % with first instars, 22 % with second, 24 % with third, and 17 % with adults. The difference was thought to be due to insect behaviour as usually more nymphs than adults settled and apparently fed on the test beans. No transmissions were obtained when parturient females or second instar males were tested. Five adults, four third instar nymphs and seven crawlers of F. virgala on test plants were compared by Posnette and Strickland (1948). The in- fection rates were 1/20 for adults, 2/20 for the third instar nymphs and 4/16 for the crawlers. Nymphs of P. njalensis removed at birth from females, colonizing CSSV 1A isolate infected seedlings, did not transmit virus when transferred to 66 test beans at a rate of 10 per bean (Dale 1958). At the Cocoa Research Institute of Ghana, nymphs for routine screening tests are from field collected females of P. njalensis which are cleaned and kept in reproduction cages. Every second day for six days nymphs are collected from the cages and the source plants infested. Thus, nymphs after the usual 48 hour AAT vary in age from 2 to 4 days when used for the lAT. Kenten (in the press) investigated the effect of age of the nymphs on transmission efficiency with batches of nymphs aged 2—3 days and 34 days after AAT. The 3—4 days old nymphs gave an infection rate of 52 % and were seemingly slightly more efficient transmitters than those aged 23 days which gave an infection rate of only 38 %. The difference was significant at P = 0.05 level. A possible reason for the difference was that the older nymphs are larger, more robust, and so less easily damaged during manipulation. The three successive batches of nymphs which were taken from the fasting adults in the reproduction cages were also compared (Kenten in the press). As the adults aged, their nymphs were less efficient as vectors. After angular transformation the infection rate of the first batch was 61.7, of the second 50.3 and of the third 42.0 (L.S.D. = 7.1, P = 0.05). However, the variation in efficiency of transmission between the nymphs of different ages and those from adults of different ages was less than the variation between different individual virus source plants and, therefore, considered of no advantage to alter the procedure outlined above for collect- ing nymphs for routine screening work. It is quite well established that there are no consistent differences in the efficiency of vectors of varying age, apart from old adult females, or males that may not feed at later developmental stages. Passage of virus into nymphs in the female reproductive organs does not seem to occur. The single report of Posnette and Robertson (1950) on differences between nymphs and adults of P. cilri, and the findings reported by Kenten (in the press) on young nymphs of P. njalensis, need to be confirmed. 288 4.8 Efficiency and virus/vector specificity of vector species Results obtained with P. citri, D. brevipes, D. sp. near brevipes, F. virgata and Pseudococcus comstocki (Kuwana) transmitting the virus CTV in Trinidad suggest that there is little if any difference in the efficiency of these five species as transmitters (Kirkpatrick 1950, 1953 b). According to Dale (1955 a), F. virgata was much inferior to P. njalensis as a vector of CSSV 1A isolate, but F. virgata seemed unsuited to the method used because of the large amount of wax produced. The insects became en- tangled within the paper cone attached to the seedlings and few settled down to feed. Consequently a fair comparison was not obtained between the species. Longworth (1964 a) allowed crawlers of P. njalensis and F. virgata to feed either on leaves or stems of CSSV infected seedlings. Infection rates were 50 % from stems and 6 % from leaves with P. njalensis but 4 % from stems and 48 % from leaves with F. virgata. A very plausible explanation for this difference was given by Entwistle and Longworth (1963) who investigated the feeding behaviour of three mealybug species. In transverse sections of cocoa stems, 164 out of 345 stylets of P. njalensis were seen to end in the phloem. Stylets of F. virgata were seen also in the phloem but less often. Phenacoccus madeiriensis Green, which is not a vector of cocoa viruses, fed differently; its stylets did not end in the phloem. The authors suggested, therefore, that mealybugs acquire virus from the phloem and must deposit virus in the phloem to infect a plant. Thus the rarity of virus transmission by F. virgata from stem may be explained by the infrequency its stylets reach the phloem. The low infection rate obtained with P. njalensis from leaves may indicate feeding preference, for Longworth (1964 a) has clearly demonstrated that F. virgata prefers the leaf as a feeding site, whereas P. njalensis, although feeding to some extent on leaves, prefers leaf axils, crevices between veins and pulvinus and scars and crevices on the stem. In the present study P. njalensis transmitted CSSV 1A isolate more effi- ciently when confined in feeding cages on stems than on leaves during the AAT. (see Table 15). Also, P. njalensis was a more efficient vector than F. virgata when both species were kept in feeding cages on stems during the AAT (see Table 27). Adegbola (1965) tested the relative efficiency of P. njalensis and F. virgata using Egbeda isolate of CSSV infected source plants and either seedlings or beans as test plants. There was a slight indication that P. njalensis is a more efficient vector when beans are used as test plants, whilst F. virgata is superior when seedlings are used as test plants. Igwegbe (1966 a) compared the relative efficiency of P. njalensis and F. virgata when virus acquisition was done on (a) leaves, (b) stems and petioles, and (c) whole seedlings. There was little difference between the two species when leaves were used as the source but otherwise P. njalensis was a more efficient vector. Adegbola (1971) did not find any difference between the transmission effi- ciency of P. njalensis and F. virgata with Egbeda isolate of CSSV. However, 289 differences caused by operator skills were obvious and P. njalensis was claimed to be more difficult to handle than F. virgata. It is clear that under identical conditions P. njalensis and F. virgata are capable to transmit virus with different efficiency. The reason for this seems to be the preference for different feeding sites and the frequency with the phloem is reached by feeding insects during the acquisition feed. The failure reported by Posnette and Robertson (1950) to obtain trans- missions in certain experiments suggested that strains of mealybugs unable to transmit might occur within the species. In tests that followed, 10 gravid females of P. njalensis were collected from widely separated points within a radius of about one mile of the laboratory of the Cocoa Research Institute, Ghana. The females and their progenies were reared separately on isolated plants in insect proof greenhouses. When tested, each of the 10 progenies trans- mitted the 1M isolate of CSSV with no striking differences in the infection rates. Lister (1953) tested colonies of P. njalensis reared from single females. The results did not provide any evidence that there are non-transmitting strains of this species. Likewise, in the present study, differences in efficiency of trans- mission were not found between four populations of P. njalensis (see Table 16). Posnette and Robertson (1950) reported that non-transmitting strains of P. citri do occur, however. Evidence of this was obtained with CSSV 1A isolate, but some strains were lost through parasitism before the tests could be concluded. Experiments with 1C isolate of CMLV showed that one strain of P. citri, which was cultured from a single female collected at Tafo, consis- tently failed to transmit under conditions suitable for transmission by the mixed progeny of five females collected at Kpeve. It was confirmed that the types were within the normal range of variation of P. citri as then recognized in West Africa. Some mealybug species such as P. njalensis, P. hargreavesi and P. citri transmit a wide range of cocoa virus isolates. Indeed, P. njalensis has trans- mitted virtually every West African isolate of cocoa viruses known to have mealybugs as vectors. In some mealybug species, however, the ability to trans- mit seems to be restricted to certain viruses or isolates as first reported by Posnette (1950) and repeatedly confirmed by many other authors. Thus, F. virgata is a more or less efficient vector of a number of isolates, but has consistently failed to transmit different CMLV isolates and 1M isolate of CSSV1) P. longispinus has failed to transmit all other isolates, except 1C of CMLV and 1M of CSSV. A detailed account of the different vectors and their trans- mission of various cocoa viruses and isolates is given in Appendices I and 11. •) . Ps. longispinus also transmitted virus IM which, except in one experiment, was not transmitted by Ferrisia virgata. When fed on plants infected with virus 1A and IM, F. virgata transmitted only IA. We do not understand why F. virgata transmitted virus 1M in one ex- periment; it is possible that a biological race was involved as with Pseudococcus citri discussed below, but no transmission was obtained when the experiment was repeated using insects from the same source. The possibility of contaminationwith another vector species cannot be eliminat- ed, but seems improbable because of the precautions taken and the distinctive appearance of Ferrisia virgata.’ (Posnette 1950). 290 4.9 Effect of physical factors on virus transmission Asomaning and Lockard (1964 a, b) and Asomaning and Kwakwa (1968) studied the effects of low and high intensity of solar radiation on development of swollen shoot disease (CSSV 1A) in cocoa seedlings. Development of virus symptoms in stems and roots was faster and more prominent in seedlings ex- posed to radiation of low intensity. In general, the seedlings showed the usual leaf symptoms of CSSV 1A infection. The low intensity seedlings produced the ’fern-leaf’ pattern while the high intensity seedlings produced the ’speckled’ pattern. Longworth (1965 b) performed similar studies with Egdeba isolate of CSSV. Bean infected cocoa seedlings were kept in darkness and in full sun- light. Germination and growth of the beans planted in full sunlight was poor and symptom recognition was confused by symptoms of scorch and mineral deficiencies. There was an indication, however, that placing the plants in the dark immediately after infection and storing them in the dark for several weeks may increase transmission rates. It is obvious that intensity of solar radiation influences development of that part of CSSV infection which is visible in symptoms, but further studies are needed to show that radiation can change the proportion of actually in- fected plants. In the present study, different combinations of light and dark during prelim- inary fasting and acquisition feed did not influence the infection rate when the lAT was carried out in the dark (see 3.10.1). Temperature had an influence upon the feeding of mealybugs and virus transmission (see 3.10.2). The effect of light on the availability of 1A isolate of CSSV from cotyledons was tested with P. njalensis (Anon. 1951 b). Fresh cocoa beans were infected by application of 30 infective crawlers per bean. After infection the embryonic axis was removed and the remaining mutilated cotyledons were used as virus source plants. These were kept in almost complete darkness or exposed to light and the availability of virus for mealybugs was tested at various intervals. In both treatments virus was first transmitted from the cotyledons seven days after infection, and virus was equally available from light and darkness treated cotyledons. The main conclusion from this result was that the essential activity of virus takes place at a level in the host cell deeper than the chloro- plast system. Certainly, symptom production appears to be a secondary effect of the virus. This is supported by the tendency even of virulent isolates to enter upon symptomless phases of infection during which transmission is still possible. 4.10 Transmission of virus complexes When Posnette and Robertson (1950) infected cocoa plants with both CSSV 1A and 1M isolates they were not mutually antagonistic. The symptoms caused by each isolate occurred independently on some leaves, but periodically leaves were formed with a characteristic ’complex’ mosaic quite unlike the symptoms caused by either isolate. To find out whether the isolates were 2917 so intimately associated in the plant that they would be simultaneously trans- mitted by single insects, four plants infected with both isolates were colonized with P. njalensis. After AAT the insects were transferred one to each test plant. One source plant gave only isolate 1A transmissions and another only isolate IM, but from two source plants both isolates were transmitted simulta- neously by single vectors. 4.11 Latent period of virus in the vector The possibility that mealybug transmitted cocoa viruses undergo a latent period in the vector was briefly discussed by Dale (1955 a). Although virus has been acquired and transmitted by P. njalensis in seven hours, composed of a four hour AAT and a three hour lAT (Posnette and Strickland 1948), or within five hours (Dale 1955 a), much longer periods are required on source plants to achieve maximum rates of transmission. In the present study a short AAT of one hour was tested with postacquisition fasting of o—4 hours and lAT of one hour, but no transmissions occurred (see 3.7.2). There is no direct evidence to conclude on the presence or absence of a latent period of virus in the mealybugs. 4.12 Type of transmission Twelve characteristics of stylet-borne (non-persistent) and circulative (persistent) viruses transmitted by homopterous insects have been selected to analyse properties of cocoa viruses with their mealybug vectors (Table 36). The data presented and discussed in this paper has been used in the analysis. However, where evaluation of data has been more difficult or definition of a characteristic is not exact, the judgement may reflect the opinion of the present author. Although the mealybug transmitted cocoa viruses have some anomalous properties, on the whole they have more characteristics of circulative viruses than of stylet-borne viruses. In view of this, it is suggested that these viruses are grouped with persistent circulative viruses rather than non-persistent stylet- borne or semi-persistent viruses. Table 36. Properties of mealybug-borne cocoa viruses according to characteristics of stylet' borne and circulative viruses. S = of cocoa viruses is characteristic of stylet-borne viruses. C = property of cocoa viruses is characteristic of circulative viruses. NK = property of cocoa viruses not known. Characteristics Stylet-borne viruses Circulative viruses Property of cocoa viruses Short transmission cycle (seconds) Long transmission cycle (hours) No latent period in vector Latent period in vector Pre-AAT fasting increases vector efficiency Pre-AAT fasting does not increase vector efficiency Short AAT (minutes) Long AAT (hours) more efficient more efficient than long than short Vector capable of infecting one or few plants Vector capable of infecting many plants Vector Vector specificity not marked specificity often marked Virus Virus non-persistent in vector persistent in vector Virus not retained in vector through a moult Virus retained in vector through a moult Virus generally transmitted by mechanical means Virus not generally transmitted by mechanical means Virus affects mainly Virus affects mainly epidermal tissue conductive tissue Virus not recoverable Virus recoverable from haemolymph from haemolymplfrom haemolymph of vector of vector Vector not infective Vector infective after inoculation of after inoculation of virus into haemocoele virus into haemocoele C Minimum cycle 5 7 hours, longer more typical NK Increases settling and S food uptake and, thus vector efficiency Minimum 1.5—2 hours, C 48—72 hours more efficient S Two plants Only mealybug vectors, C specificity of F. virgata and P. longispinus Maximum persistence 72 —96 C hours, 50 % persistence about 12 hours C Mechanical transmission C occurs after chemical processing C NK NK more efficient conductive tissue 292 293 Summary Several experiments concerning mealybugs (Pseudococcidae) and their transmission of cocoa viruses are described and discussed in this paper. P. nja- lensis, P. cilri and F. virgata reacted positively towards low intensity of light in 25—27° C. and 40—60 % RH. In the same temperature the three mealybug species and P. longispinus reacted more positively towards high than to low humidities. P. njalensis reacted more positively towards 20—25° C. than higher or lower temperatures. The most favourable temperature for reproduc- tion of fasting females of P. njalensis was 25—33° C. with 55 —75 % RH. Partial fecundity of field collected females was then 92 nymphs per female. Biological half-life of 32 P in feeding P. njalensis was 11.4 or 15.3 days de- pending on external conditions. The amount of food uptake, measured in 32P- activity from virus infected plants was positively related to the infectivity status of P. njalensis with CSSV IA. Plant radioactivity was positively related to mealybug radioactivity. There was a positive correlation between the amount of honeydew excreted by P. njalensis and the radioactivity of the honeydew. Frequency distribution of radioactivity in P. njalensis after feeding on 32P- active seedlings or sucrose solution was closely similar. Radiation effects on infectivity of P. njalensis were not detected at a level of 30 /iCi/ml of 32 P in seedling culture solution. With the virus CSSV 1A and P. njalensis as the vector, preliminary fasting increased the rate of settling down to feed, the amount of food uptake and the infectivity of mealybugs. Virus was transmitted more frequently from the stems than the leaves of infected plants. Groups of young seedlings of 29 —54 days were almost equally good sources of virus for mealybugs, and availability of virus varied significantly between source plants of the same age in closely similar conditions. Mean duration of stylet withdrawal from source plants was 240 seconds. The optimum length of acquisition access time for maximum rate of transmission with prestarved mealybugs was 48—72 hours, with shorter or longer acquisition access time the infection rate was lower. Virus was not acquired by mealybugs during short probes on source plants or short feeds of one hour maximum. Virus was readily acquired from infected seedlings via a parafilm membrane by mealybugs as indicated by subsequent transmission into test plants. Acquisition of virus from liquid leaf extract via a parafilm membrane and subsequent transmission was rare. After acquisition feed the virus persisted in the mealybugs up to 72—96 hours and loss of infectivity was exponential with time. The half-life of persistence was 12 —l3 hours. Virus persisted in the mealybugs through postacquisition moulting. Postacquisition feeding increased virus persistence. Latent period of virus in the test plants 294 was short and normally over 98 % of the infected plants sowed symptoms of virus in three months. With different numbers of mealybugs on test plants during the inoculation feed, the infection rate was according to laws of probab- ility. The use of mealybugs while testing virus resistant and tolerant cocoa types is demonstrated. No evidence was obtained that location of feeding mealybugs on test beans influences the rate of virus transmission. Virus trans- mission by mealybugs was not affected by light or dark. The amount of food uptake by mealybugs was higher and also virus was transmitted more frequently in 29 36° C. than in lower temperatures. Although investigated to a lesser extent, the transmission characteristics of CSSV 1M or CMLV 1C with P. njalensis as the vector, or CSSV 1A with F. virgata, were closely similar to those of CSSV 1A with P. njalensis as the vector. A general mathematical formula is given to describe the increase of infec- tivity in mealybugs during acquisition feed on basis of linear food uptake and exponential loss of infectivity. 295 REFERENCES Adegbola, M. O. K, 1965. The relative efficiency of Planococcoides (= Pseudococcus) njalensts (Laing) and Ferrisiana virgala (Ckll) in the transmission of CSSV. Ann. Rep. Cocoa Res. Inst. Nigeria 1963 64, p. 43 44. » 1971. The efficiency of mealybug handlers as a factor in laboratory transmission of cocoa swollen shoot virus (CSSV). Niger. Ent. Mag. 2: 75 82. Anon. 1950. Virus-host relationships, (a) A newly-discovered locus of infection in the plant. Ann. Rep. W. Afr. Cacao Res. Inst. 1948—49, p. 13 14. » 1951 a. Laboratory rearing of mealybugs. Ann. Rep. W. Afr. Cacao Res. Inst. 1949 50, p. 65-66. » 1951 b. Virus-host relationships. Ann. Rep. W. Afr. Cacao Res. Inst. 1949 50, p. 13 19. » 1953 a. Rearing of mealybugs for parasite breeding. Ann. Rep. W. Afr. Cacao Res. Inst. 1950-51, p. 41-42. » 1953 b. Quart. Rep. W. Afr. Cacao Res. Inst. 32: 7 (mimeogr.). » 1955. Quart. Rep. W. Afr. Cocoa Res. Inst. 39: 16 (mimeogr.). » 1957. Quart. Rep. W. Afr. Cocoa Res. Inst. 47:6 (mimeogr.). » 1958. Quart. Rep. W. Afr. Cocoa Res. Inst. 52: 5 6 (mimeogr.). * 1959. Quart. Rep. W. Afr. Cocoa Res. Inst. 56: 7 (mimeogr.). * 1963. Half-Yearly Rep. Cocoa Res. Inst. Ghana 2: 13 (mimeogr.). Asomaning, E. J, A. & Kwakwa, R. S. 1968. Effects of degrees of shading on the growth of virus-infected seedlings. Ann. Rep. Cocoa Res. Inst. Ghana 1965 66, p. 114 125. » & Lockard, R. G. 1964 a. The effects of solar radiation on the growth of cocoa infected with New Juaben virus. Ann. Rep. Cocoa Res. Inst. Ghana 1962—63. p. 88. —» 1964 b. Studies on the physiology of cocoa (Theobroma cacao L.). I. Suppression of swollen-shoot virus symptoms by light. Ann. Appi. Biol. 54: 193 198. Attafuah, A. & Brunt, A. A. 1960. Classification of isolates. Ann. Rep. W. Afr. Cocoa Res. Inst. 1958-59, p. 16-17. Baker, R. E. D. & Dale, W. T. 1947. Notes on a virus disease of cacao. Ann. Appi. Biol. 34: 60-65. Banks, C. J. & Nixon, H. L. 1959. The feeding and excretion rates of Aphis fabae Scop, on Vida faba L. Ent. Exp. Appi. 2:77 81. Blencowe, J. W. 1961. A virus disease of cocoa in Sierra Leone. Ann. Rep. W. Afr. Cocoa Res. Inst. 1959-60, p. 20. » 1962. Future plans for cocoa breeding in Ghana. Rep. Cocoa Conf. London 1961, p. 141-145. —» & Attafuah, A. 1959. Resistance and tolerance. Ann. Rep. W. Afr. Cocoa Res. Inst. 1957-58, p. 24-25. * , Brunt, A. A., Kenten, R. H. & Lovi, N. K. 1963. A new virus disease of cocoa in Sierra Leone. Trop. Agric. (Trinidad) 40: 233 236. Box, H. M. 1945. Insect transmission of the ’swollen-shoot’ virus in West Africa. Nature (Lon- don) 155:608-609. Brunt, A. A. & Kenten, R. H. 1960. Mechanical transmission of cocoa swollen-shoot virus. Virology 12: 328-330. * 1962. Mechanical transmission of cocoa swollen-shoot virus to and from cocoa and other hosts. Ann. Appi. Biol. 50: 749 754. Carter, W. 1956. Notes on some mealybugs (Coccidae) of economic importance in Ceylon. FAO Pi. Protect. Bull. 4: 49 52. 296 Carter, W. 1961. Report to the International Cooperation Administration and USOM/Ghana. 31 p. (mimeogr. available at Cocoa Res. Inst. Ghana). » 1963. Mealybug wilt of pine-apple; a reappraisal. Ann. N. Y. Acad. Sci. 105: 741 764. Chase, G. D. & Rabinowitz, J. L. 1962. Principles of radioisotope methodology. 372 p. Minnea- polis. Cornwell, P. B. 1957. A preliminary investigation into the uptake and translocation of phos- phorus-32 in Theobroma cacao L. Trop. Agric. (Trinidad) 34: 117 132. Cotterell, G. S. 1943. Swollen-shoot virus. Rep. Centr. Cocoa Res. Sta. Tafo 1938 42, p. 51-55. Dale, W. T. 1954a. Quart. Rep. W. Afr. Cacao Res. Inst. 33: 7 (mimeogr.) * 1954 b. Quart. Rep. W. Afr. Cacao Res. Inst. 34: 6 (mimeogr.). » 1954 c. Quart. Rep. W. Afr. Cacao Res. Inst. 35:9 (mimeogr.). » 1954 d. Quart. Rep. W. Afr. Cacao Res. Inst. 36: 5 (mimeogr.). —» 1955 a. Optimum conditions for transmission. Ann. Rep. W. Afr. Cocoa Res. Inst. 1954-55, p. 33-35. » 1955 b. Quart. Rep. \V. Afr. Cacao Res. Inst. 37: 5—6 (mimeogr.). » 1957. Insect transmission studies. Ann. Rep. W. Afr. Cocoa Res. Inst. 1955 56, p. 30-31. » 1958. Insect-transmission studies. Ann. Rep. W. Afr. Cocoa Res. Inst. 1956—57, p 22-24. » 1962. Diseases and pests of cocoa. A. Virus diseases. Agriculture and Land Use in Ghana, p. 286 316. London. Day, M. F. & Irzykiewicz, H. 1953. Feeding behaviour of the aphids Myzus persicae and Brevicoryne brassicae, studied with radiophosphorus. Austr. J. Biol. Sci. 6:98 108. » & McKinnon, A. 1951. A study of some aspects of the feeding of the jassid Orosius. J. Sci. Res. Ser. B 4: 125-135. Decker, F. E. 1956. Quart. Rep. W. Afr. Cocoa Res. Inst. 41; 8 (mimeogr.). —» 1957. Mass-breeding of mealybugs. Ann. Rep. W. Afr. Cocoa Res. Inst. 1955 56, p. 41-42. Donald, R. G. 1954a. Quart. Rep. W. Afr. Cacao Res. Inst. 33: 7 (mimeogr.). » 1954 b. Quart. Rep. W. Afr. Cacao Res. Inst. 35: 18 (mimeogr.). » 1954 c. Quart. Rep. W. Afr. Cacao Res. Inst. 36: 20 (mimeogr.). » 1955. Quart. Rep. W. Afr. Cacao Res. Inst. 37; 21 (mimeogr.). » 1957. Mass-breeding of mealybugs. Ann. Rep. W. Afr. Cocoa Res. Inst. 1955 56, p. 41-42. Duffus, J. E. & Gold, A. H. 1967. Relationship of tracer-measured aphid feeding to acquisition of beet western yellows virus and to feeding inhibitors in plant extracts. Phytopath. 57: 1237-1241. Eguagie, W. E. 1970. Non-transmittability of Offa-Igbo isolate of cocoa swollen shoot virus (CSSV) by Aphis gossypii Glover (Homoptera: Aphididae). Niger. Ent. Mag. 2:49 55. Elmer, C. H. 1922. Mosaic cross-inoculation and insect transmission studies. Science 56: 370. » 1925. Transmissibility and pathological effects of the mosaic diasease. Res. Bull. lowa Agric. Exp. Sta. 82: 39. Entwistle, P. F. & Longworth, J. F. 1963. The relationship between cacao viruses and their vectors: the feeding behaviour of three mealybug (Homoptera: Pseudococcidae) species. Ann. Appi. Biol. 52: 387-391. Fajardo, T. C. 1930. Studies of the mosaic disease of the bean (Phaseolus vulgaris). Phytopath. 20: 469. Gibbs, A. J. & Gower. J. C. 1960. The use of a multiple-transfer method in plant virus trans- mission studies some statistical points arising in the analysis of results. Ann. Appi. Biol. 48: 75 83. Hafez, M, & Salama, H. S. 1970. Biological studies on the sugar cane mealy bug, Sacchari- coccus sacchari Ckll., in Egypt. Bull. Soc. Ent. Egypte 53: 499—516. * , Salama, H. S. & Amin, A. H. 1967. Population dynamics of the plum scale. Paria- toria oleae (Colvee), in Egypt (Coccidae). Bull. Soc. Ent. Egypte 51: 347 357. Hamilton, M. H. 1935. Further experiments on the artificial feeding of Myzus persicae (Sulz.). Ann. Appi. Biol. 22:243-258. 297 Hughes, W. A. & Lister, C. A. 1953. Lime dieback in the Gold Coast, a virus disease of the lime. Citrus aurantifolia (Christerum) Swingle. J. Hort. Sci. 28: 131. Igwegbe, E. C. K. 1966a. The preferences shown by Planococcoides njalensis (Laing) and Ferri- siana virgata (Ckll.) for different feeding sites and its effect on their efficiency as vectors. Ann. Rep. Cocoa Res. Inst. Nigeria 1964—65, p. 60 62. * 1966 b. Availability of C.S.S.V. in different cocoa clones to Planococcoides njalensis (Laing). Ann. Rep. Cocoa Res. Inst. Nigeria 1964 65, p. 62 65. Imms, A. D. 1957. A general textbook of entomology. 886 p. London. James, H. C. 1937. Sex ratios and the status of the male in Pseudococcinae (Hem. Coccidae). Bull. Ent. Res. 28:429-461. Kenten, R. H. Factors affecting the transmission of virus by mealybugs. Ann. Rep. Cocoa Res. Inst. Ghana 1972 —73 (in the press). —» & Legg, J. T. 1967. Some properties of cocoa mottle leaf virus. J. Gen. Virol. 1: 465 470. » 1970. Methods of assessing the tolerance and resistance of different types of cocoa to cocoa swollen-shoot virus. Ann. Appi. Biol. 65:419—424. » 1971. Serologicalrelationships of some viruses from cocoa (Theobroma cacao L.) in Ghana. Ann. Appi. Biol. 67:195-200. » & Owusu, G. K. 1970. Cocoa necrosis virus. Ann. Rep. Cocoa Res. Inst. Ghana 1968 69, p. 35. Kirkpatrick, T. W. 1950. Insect transmission of cacao virus disease in Trinidad. Bull. Ent. Res, 41:99-117. » 1953 a. Insect pests of cacao and insect vectors of cacao virus disease. Rep. Cacao Res. Imp. Coll. Trop. Agric. Trin. 1945 51, p. 122 125. » 1953 b. Notes on minor insect pests of cacao in Trinidad. Rep. Cacao Res. Imp. Coll. Trop. Agric. Trin. 1952, p. 62 71. Knight, R. & Tinsley, T. W. 1958. Some histological observations on virus-infected Theobro- broma cacao L. Ann. Appi. Biol. 46: 7 10. Legg, J. T. & Bonney, J. K. 1967. The host range and vector species of viruses from Cola chlamydanlha K. Schum., Adansonia digilata L. and Theobroma cacao L. Ann. Appi. Biol. 60:399-403. —» 1968 a. Insect transmission studies. Ann. Rep. Cocoa Res. Inst. Ghana 1965 66, p. 30-31. » 1968 b. Availability of CSSV in sensitive and tolerant varieties of cocoa. Ann. Rep. Cocoa Res. Inst. Ghana 1966 67, p. 27—28. —» & Kenten, R. H. 1970. The resistance and tolerance of different cocoa varieties to cocoa swollen-shoot virus in Ghana. Ann. Appi. Biol. 65:425 434. Lister, R. M. 1953. Ann. Rep. W. Afr. Cacao Res. Inst. 1952 53, p. 9 10. —» & Thresh, J. H. 1954. Mealybug vectors. Ann. Rep. W. Afr. Cacao Res. Inst. 1953 54, p. 13. Lockard, R. G. & Asomaning, E. J. A. 1963. Composition of complete nutrient solutions. Ann. Rep. W. Afr. Cocoa Res. Inst, 1961 62, p. 56 57. Lodos, N. & Boafo, E. O. 1968. Studies of ants associated with mealybugs and virus vector investigations. Ann. Rep. Cocoa Res. Inst. Ghana 1966—67, p. 38. Longworth, J. F. 1964a. The preference shown by P. njalensis and F. virgata for different feeding sites, and its effect on their efficiency as vectors. Ann. Rep. Cocoa Res. Inst. Nigeria 1962-63, p. 29-30. » 1964 b. Availability of CSSV in different cocoa types. Ann. Rep. Cocoa Res. Inst. Ni- geria 1962-63, p. 31-33. » 1964 c. Half-Yearly Rep. W. Afr. Cocoa Res. Inst. (Nigeria) 2: 20—21 (mimeogr.). » 1965 a. Ann. Rep. W. Afr. Cocoa Res. Inst. (Nigeria) 1963—64, p. 34. » 1965 b. The effect of light on transmission of cocoa swollen shoot virus by insects and mechanical methods. Ann. Rep. Cocoa Res. Inst. Nigeria 1963—64, p. 46. » & Entwistle, P. F. 1965. Virus/vector relationships. Ann. Rep. Cocoa Res. Inst. Nigeria 1963- 64. p. 40-43. Madge, D. S. 1961. The control of relative humidity with aqueous solutions of sodium hydroxide. Ent. Exp. Appi. 4; 143 147. 298 Markkula, M. 1963. Studies on the pea aphid Acyrthosiphon pisum Harris (Horn., Aphididae). with special reference to the differences in the biology of the green and red forms. Ann. Agric. Fenn. 2, Suppl. 1: 1—3O p. Martini, C. K. H. 1959 a. Quart. Rep. W. Afr. Cocoa Res. Inst. 55: 12—l3 (mimeogr.). » 1959 b. Quart. Rep. W. Afr. Cocoa Res. Inst. 56: 16 (mimeogr.). » 1960. Insect transmission studies. Ann. Rep. W. Afr. Cocoa Res. Inst. 1958—59, p. 67-68. » 1961. Transmission of Nigerian cocoa viruses. Ann. Rep. W. Afr. Cocoa Res. Inst. 1959 60, p. 68-69. —» 1962. Vectors of Nigerian cocoa viruses. Ann. Rep. W. Afr. Cocoa Res. Inst. 1960—61, p. 75-76. Mittler, T. E. & Dadd, R. H. 1954. An improved method for feeding aphids on artificial diets. Ann. Ent. Soc. Amer. 57; 139 140. Newton, W. 1953. Transmission of tobacco mosaic by citrus mealybug. FAO PI. Protect. Bull. 2; 40. Nicol, J. 1950. Quart. Rep. W. Afr. Cacao Res, Inst. 25:4 (mimeogr.). Okusanya, B. A. M. 1968. Studies of cacao swollen shoot disease. Ann. Rep. Cocoa Res. Inst. Nigeria 1966 67, p. 66 69. * 1969. Cocoa swollen shoot virus. Ann. Rep. Cocoa Res. Inst. Nigeria 1967 68, p. 88 92. » 1970. Transmission of CSSV. Ann. Rep. Cocoa Res. Inst. Nigeria 1968—69, p. 55 56. » 1971. Breeding of Pseudococcus njalensis. Ann. Rep. Cocoa Res. Inst. Nigeria 1969 70, p. 46-47. Olitsky, P. K. 1925. The transfer of tobacco and tomato mosaic diseases by Pseudococcus citri. Science 62: 442. Owusu, G. K. 1959. The availability of swollen shoot virus to mealybug vectors in tolerant and sensitive cocoa. Ann. Rep. Cocoa Res. Inst. Ghana 1967 68, p. 32 34. » 1970. Availability of cocoa swollen shoot virus to mealybugs from tolerant and sensitive cocoa. Ann. Rep. Cocoa Res. Inst. Ghana 1968 69, p. 31 33. » 1971 a. Cocoa necrosis virus in Ghana. Trop. Agric. (Trinidad) 48: 133—139. » 1971 b. Availability of cocoa swollen shoot virus to mealybugs from tolerant and sensi- tive cocoa. Proc. 3rd Intern. Cocoa Res. Conf. 1969, p. 512—517. » 1972. Acquisition of swollen shoot virus by mealybugs from cocoa plants during the pe- riod of latent infection. Ann. Rep. Cocoa Res. Inst. Ghana 1969 70, p. 60 61. » 1973. Availability of cocoa swollen shoot virus to mealybugs from tolerant and sensitive cocoa. Ann. Rep. Cocoa Res, Inst. Ghana 1970 71, p. 63. —» & Bonney, J. K. 1972. Studies of CSSV isolates from Ashanti region. Ann. Rep. Cocoa Res. Inst. Ghana 1969 70, p. 63 65. » & Kenten, R. H. 1972. Cocoa necrosis virus. Ann. Rep. Cocoa Res. Inst. Ghana 1969 70, p. 61-63. Posnette, A. F. 1940. Transmission of swollen-shoot. Trop. Agric. (Trinidad) 17: 98. » 1941. Swollen-shoot virus disease of cacao. Trop. Agric. (Trinidad) 18: 87 89. » 1944. Virus diseases of cacao in Trinidad. Trop. Agric. (Trinidad) 21: 105 106. » 1947 a. Virus diseases of cacao in West Africa. I. Cacao viruses lA, 18, 1C and ID. Ann. Appi. Biol. 34:388 402. * 1947 b. The use of seeds in the insect transmission of some plant viruses. Nature (London) 159:500. —» 1950. Virus diseass of cacao in West Africa VII. Virus transmission by different vector species. Ann. Appi. Biol. 37: 378 384. » 1951. Virus research at the West African Cacao Research Institute, Tafo, Gold Coast. Trop. Agric. (Trinidad) 28: 133 142. —» & Robertson, N. F. 1950. Virus diseases of cacao in West Africa. VI. Vector inves- tigations. Ann. Appi. Biol. 37:363 377. —* & Strickland, A. H. 1948. 111. Technique of insect transmission. Ann. Appi. Biol. 35: 53-63. » & Todd, J. McA. 1951. VIII. The search for virus resistant cacao. Ann. Appi. Biol. 38: 785-800. » 1955. IX. Strain variation and interference in virus IA. Ann. Appi. Biol. 43:433 453. 299 Roivainen, O. 1968. Insect transmission studies. Ann. Rep. Cocoa Res. Inst. Ghana 1966 —67, p. 28-30. » 1969. Effect of vector feeding on virus transmission. Ann. Rep. Cocoa Res. Inst. Ghana 1967 68, p. 40. » 1970. Virus/vector relationships. Ann. Rep. Cocoa Res. Inst. Ghana 1968—69, p. 29—30. » 1971. Circulative transmission of cocoa swollen shoot virus by the mealybug Planococcoi- des njalensis (Laing) (Homoptera: Pseudococcidae). Proc. 3rd Intern. Cocoa Res. Conf. 1969, p. 518-521. » 1972 a. Mechanism of transmission. Ann. Rep. Cocoa Res. Inst. Ghana 1969 70, p. 59-60. » 1972 b. Optimum conditions for transmission, Ann. Rep. Cocoa Res. Inst, Ghana 1969 70, p. 57-59. » 1973. Viruses and Coccoidea. Front. Biol. 31:455—462. Amsterdam. Salama, H. S. 1970a. Reactions of the grape mealy bug, Planococcus vitis Nied., towards some environmental factors. Bull. Soc. Ent. Egypte 53; 271 281. * 1970 b. Population dynamics of the scale insect Mycetaspis personatus (Comstock) in Egypt (Homoptera-Coccoidea) . Z. Angew. Ent. 66:42—46. Semangun, H. 1961. Mosaic symptoms on cocoa leaves in Java. Phytopath. Lab. Fac. Agr. For. Univ. Gadjahmada, Jogjakarta, Indonesia 2: 1—9 p. Seuge, J., Morere, J. L. & Ferradini, C. 1971. Effect of food preirradiation on the fecundity of two insects: Mealy bugs (Pseudaulacaspis pentagona Targ.) and Indian-raeal moths (Plodia interpunctella Hiibn.). Radiat. Res. 45:210 215. Snedecor, G. W. & Cochran, W. G. 1968. Statistical methods. 593 p. Ames. Solomon, M. E. 1945. The use of cobalt salts as indicators of humidity and moisture. Ann. Appi. Biol. 32:75-85. Steven, W. F. 1936. A new disease of cocoa in the Gold Coast. Gold Coast Farm. 5: 122. (reprint- ed Trop. Agric. (Trinidad) 14: 84). Strickland, A. H. 1951. The entomology of swollen shoot of cacao. I. The insect species involved, with notes on their biology. Bull. Ent. Res. 41: 725 748. Sylvester, E. S. 1958. Aphid transmission of plant viruses. Proc. 10th Intern. Congr. Ent. 1956, 3: 195-200. Teliz, D., Grogan, R. G. & Lownsberry, B. F. 1966. Transmission of tomato ringspot, peach yellows bud mosaic, and grape yellow vein viruses by Xiphinema americanum. Phy- topath. 56: 658 663. Thorold, C. A. 1975. Diseases of cocoa. 423 p. London. Thresh, J. M, 1958 a. The control of cacao swollen shoot disease in West Africa. Techn. Bull. W. Afr. Cocoa Res. Inst. 4: 1—36 p. * 1958 b. The spread of virus disease in cacao. Techn. Bull. W. Afr. Cocoa Res. Inst. 5: 1-36 p. » 1958 c. The availability of cacao swollen-shoot virus to mealybugs feeding on infected trees. Ann. Rep. W. Afr. Cocoa Res. Inst. 1956 57, p. 78 81. —» & Tinsley, T. W. 1959. The viruses of cacao. Techn. Bull. W. Afr. Cocoa Res. Inst. 7: 1-32 p. Youdeowei, A. 1967. The reactions of Dysdercus intermedins (Heteroptera, Pyrrhocoridae) to moisture, with special reference to aggregation. Ent. Exp. Appi. 10: 194 210. —* 1968. The behaviour of a cotton Stainer Dysdercus intermedins (Heteroptera, Pyrr- hocoridae) in a temperature gradient and the effect of temperature on aggregation. Ent. Exp. Appi. 11:68 80. Watson, M. A. & Nixon, H. L, 1953. Studies on the feeding of Myzus persicae (Sulz.) on ra- dioactive plants. Ann. Appi. Biol. 40; 537 545. Welsh, M. F. 1961. Terminology of plant virus diseases. Can. J. Bot. 39; 1773 l7BO. 300 Selostus Villakirvat (Homoptera: Pseudococcidae) kaakaopuun virusten siirtäjinä Osmo Roivainen Maatalouden tutkimuskeskus, Tuhoeläintutkimuslaitos, 01301 Vantaa 30 Tässä julkaisussa selostetaan Ghanassa 1966 —7l suoritettuja tutkimuksia kaakaopuun viruksista ja niiden villakirvavektoreista (Pseudococcidae). Planococcoides njalensis, Plano- coccus citri ja Ferrisia virgata osoittivat käyttäytymisellään, että niillä on positiivinen valoreak- tio. Samat kolme villakirvalajia ja Pseudococcus longispinus hakeutuivat n, 90 %:n suhteelliseen kosteuteen jakarttoivat kuivia olosuhteita. P. njalensis-lajin lämpöpreferenssialue oli 20—25° C., mutta suotuisin lisääntymislämpötila oli 25 33° C. suhteellisen kosteuden ollessa 55 —75 %. Radiofosforin biologinen puoliintumisaika P. njalensis- lajissa oli 11.4 15.3 päivää. Villa- kirvojen viruksen saastuttamista kasveista ottaman ravinnon määrän ja viruksensiirtokyvyn välillä oli positiivinen korrelaatio. Ravintokasvin radioaktiivisuutta lisättäessä myös villakirvo- jen radioaktiivisuus lisääntyi samassa suhteessa. Villakirvojen erittämän mesikasteen spesifinen radioaktiivisuus vaihteli vähän. Radioaktiivisen fosforin frekvenssijakautuma oli samanlainen villakirvoissa, jotka olivat syöneet radioaktiivista kasvia tai sakkaroosiliuosta. Villakirvojen kyky siirtää virusta ei huonontunut radioaktiivisesta fosforista, jotakäytettiin 30 ij.Cijml kasvien ravintoliuoksessa. Tutkittaessa kaakaovirusta CSSV IA ja P. njalensis-lajia vektorina, syömään asettuminen nopeutui, ravinnonotto lisääntyi ja viruksensiirtokyky parani kun vektoreita paastotettiin ennen saamasyöntiä. Vektorit siirsivät virusta tehokkaammin kaakaopuun varsista kuin lehdistä. Nuo- ret 29 54 päivää vanhatkasviryhmät olivat vektoreille yhtä hyviä viruslähteitä, mutta yksittäi- set kasvit erosivat toisistaan tässä suhteessa suurestikin. Vektorilta kului aikaa keskimäärin 240 sekuntia suuosien pois vetämiseen kasvisolukosta. Vektorit siirsivät virusta tehokkaimmin kun saama-aika oli 48—72 tuntia. Kun saama-aika oli yksi tunti tai lyhyempi vektorit eivät siirtä- neet virusta. Lehtien mehusta valmistetusta uutteesta vektorit siirsivät virusta vain harvoin. Virus oli vektorissaan puolipysyvä ja 72 —96 tunnin kuluttua saamasyönnistä saastunta oli mil- tei kokonaan hävinnyt. Saastunnan häviäminen oli eksponentiaalista luonteeltaan. Häviämisen puoliintumisajaksi saatiin 12—l3 tuntia. Saama-ajan jälkeinen nahanluonti, paastoaminen tai syöminen eivät poistaneet vektorin viruksensiirtokykyä. Viruksen piilovaihe kasveissa oli lyhyt ja noin kolmen kuukauden kuluttua inokulaatiosyönnistäyli 98 % virustartunnan saaneis- ta testikasveista osoitti taudin oireita. Lisättäessä vektorien lukumäärää inokulaatiosyönnin aikana, viruksensiirtotehokkuus kasvoi todennäköisyysoppien mukaan. Valo tai pimeys eivät vaikuttaneet siirtotehokkuuteen, sen sijaan lämpötilalla oli merkitystä. Kahdella muulla kaakao- viruksella tai isolaatilla ja F. virgata-la.ji\la. saatiin edellä selostetun kaltaisia tuloksia. Tutkimuksessa pohditaan myös villakirvojen soveltuvuutta kaakaolajikkeiden virusresis- tenssin mittaamiseen. Lisäksi on johdettu siirtämistehokkuuden muutoksia kuvaava mate- maattinen malli jota on testattu kokeellisesti. Appendix I. Vectors of cocoa viruses (with some mealybug species synonyms that most often appear in cocoa virus literature) Abbreviations: CMLV = cocoa mottle leaf virus CTV = cocoa Trinidad virus CSSV = cocoa swollen shoot virus CCV = Ceylon cocoa virus Vector species Virus Reference 1. Delococcus tafoensis (Strickland) CSSV Anon. 1958 syn. Formicoccus tafoensis Strickland 2. Dysmicoccus brevipes (Cockerell) CMLV Posnette 1950 CSSV Posnette 1950 CTV Kirkpatrick 1950 3. Dysmicoccus sp. near brevipes (Cockerell) CTV Kirkpatrick 1950 4. Ferrisia virgata (Cockerell) CSSV Posnette and Strickland 1948 CTV Kirkpatrick 1950 5. Maconellicoccus ugandae (Laing) CSSV Posnette 1950 syn. Phenacoccus sp. (H 6418 in Tafo Collection) 6. Paracoccus sp. near proteae (Hall) CSSV Lister 1953 syn. Pseudococcus sp. near proteae 7. Paraputo anomalus (Newstead) CSSV Posnette 1950 syn. Paraputo ritchiei Laing 8. Planococcoides njalensis (Laing) CMLV Posnette 1950 CSSV Posnette and Strickland 1948 9. Planococcus sp. near eettis (Strickland) CMLV Posnette 1950 syn. Pseudococcus sp. near eettis Strickland CSSV Posnette 1950 10. Planococcus citri (Risso) CMLV Posnette 1950 CSSV Posnette 1950 CTV Kirkpatrick 1950 CCV Carter 1956 11. Planococcus kenyae (Le Pelley) CMLV Dale 1957 CSSV Dale 1957 12. Planococcus lilacinus (Cockerell) CCV Carter 1956 13. Pseudococcus comstocki Kuwana CTV Kirkpatrick 1953 a 14. Pseudococcus concavocerarii James CMLV Attafuah and Brunt 1960 CSSV Posnette 1950 15. Pseudococcus sp. near fragilis Brain CMLV Attafuah and Brunt 1960 syn. Pseudococcus sp. near gahani Green CSSV Posnette 1950 16. Pseudococcus hargreavesi Laing CMLV Posnette 1950 syn. Pseudococcus bukobensis (Laing) CSSV Posnette 1950 17. Pseudococcus longispinus (Targioni Tozzetti) CSSV Posnette 1950 syn. Pseudococcus adonidum (L.) 18. Pseudococcus sp. near masakensis James CSSV Posnette 1950 19. Tylococcus westwoodi Strickland CSSV Anon. 1953 b, Lister and Thresh 1954 301 Appendix 11. Cocoa virus isolates and their vectors. 4- = transmission, = no transmission Virus, country Vector species (see Appendix I for key of numbers) "R pfprpnrp and isolate 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 «eierence CSSV, Ghana Aboboya + + 4- + + + Legg and Bonney 1967 Acherechere + Anon. 1963 Adiembra + 4- + + + 4- Legg and Bonney 1967 Aiyiboso 4" Dale 1958 Amafie 4- Dale 1958 + 4- 4- 4- 4- +4- Attafuah and Brunt 1960 Amakom 4- Owusu and Bonney 1972 Anibil 4- 4- 4- 4- 4- 4- Legg and Bonney 1967 Ankra-Nkwanta 4" Dale 1958 Asuboi Cement + Anon. 1953 Bisa 4- 4- Posnette 1947 a + 4- 4- 4- 4- Attafuah and Brunt 1960 Bobiriso 4- Owusu and Bonney 1972 Bosomtwe 4-4- 4- Posnette 1950 -f +4-4-4- + + + Attafuah and Brunt 1960 Bosomuoso + Dale 1958 + + + + + + + + Attafuah and Brunt 1960 Dawa + + + + + Posnette 1950 Enchi + + Anon. 1963 Koben + Owusu and Bonney 1972 Konongo + + + + + Posnette 1950 Mampong (1M) + + + + + Posnette 1950 + + + + ++ + Attafuah and Brunt 1960 Morso + Owusu and Bonney 1972 New Juaben (1A) + + + + + + + Posnette 1950 + Dale 1955a 4- Dale 1957 + Anon. 1958 Dale 1958 Anon. 1959 Nkawkaw + + Posnette 1950 + + + + + + + + Attafuah and Brunt 1960 Nsaba + Anon. 1957 Onyimso + Owusu and Bonney 1972 Osino + Dale 1958 Peki + + Dale 1958 Sedi-Nkawie + Dale 1958 Wiawso ++ + + + + + + Posnette 1950 Worawora + Dale 1958 CSSF, Nigeria Akanran (Egbeda) + + + + Posnette 1950 Balogun + + Thresh 1) Ife + Okusanya 1968 Ikire + + Thresh 1) Hare + + Thresh 1) Ilesha + + + + + Posnette 1950 Klepe + + Thresh 1) *) Communication by letter June 1976 302 Appendix 11. Vector species (see Appendix I for key of numbers) (continued) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 Reference Offa-Igbo + + + 4- + 4- Posnette 1950 4- Anon. 1953 b 4- 4-4-4-4- 4-4-4- Attafuah and Brunt 1960 Olanla 4- 4-4-4- 4- Posnette 1950 Ondo 4- 4- Longworth 1965 a CSSP, Ivory Coast Kongodia 4- 4-4-4- Posnette 1950 Yakasse 4-4-4- Posnette 1950 4- Anon. 1958 4- 4-4- 4- Attafuah and Brunt 1960 CSSV, Sierra Leone Gandorhun 4- 4-4-4- 4- Blencowe 1961 4-4-4- 4- 4- Attafuah and Brunt 1960 CML V, Ghana Kpeve (1C) 4- 4-4-4- 4- Posnette 1950 Anon. 1959 4- 4- 4- 4- 4- 4- 4- Attafuah and Brunt 1960 AD 7 -4-4- Dale 1958 4- 4- 4- 4- 4- Attafuah and Brunt 1960 4- 4- 4- 4- 4- Legg and Bonney 1967 AD 14 - 4-4- Dale 1958 Legg and Bonney 1967 AD 36 - Dale 1958 4- 4- 4- 4- 4- Legg and Bonney 1967 AD 75 (Wasipe) 4- 4- Dale 1958 4- 4- 4- 4- -)- Legg and Bonney 1967 AD 191 (Wusuta) - 4-4- Dale 1958 Legg and Bonney 1967 AD 196 Legg and Bonney 1967 CMLV, Nigeria Alaparun 4-4- 4- Posnette 1950 4-4-4--)- 4-4-4- Attafuah and Brunt 1960 CTV, Trinidad 4-4-4- 4-4- Kirkpatrick 1950, 1953 a CCV, Sri Lanka 4-4- Carter 1956 CN V, Ghana Sesamang Nkoranza Kenten and Owusu 1970 Owusu and Kenten 1972 CNV, Nigeria Asalu Posnette 1950 - - Attafuah and Brunt 1960 Martini 1960 CYMV, Sierra Leone Giehun Blencowe et al. 1963 Mosaic disease, Java details of vector tests not known Semangun 1961 303 304 Appendix 111. Insect species that have failed to transmit cocoa viruses. Insect species Virus Reference Aphis gossypii Glover1) CSSV Martini 1961, 1962, Longworth 1964 c, CNV Eguagie 1970 Mesohomotoma tesmanni Aulman CSSV Box 1945, Posnette and Strickland CNV 1948, Lister 1953, Martini 1960 Toxoptera aurantii B.d.F, CSSV Box 1945, Posnette and Strickland CNV 1948, Lister 1953, Martini 1960, Carter 1961 Toxoptera citricidus Kirkaldy CNV Martini 1960 Geococcus coffeae Green CSSV Posnette 1951, Anon. 1955 Geococcus sp. CSSV Lodos and Boaeo 1968 Orthezia insignis Browne CTV Kirkpatrick 1950 Orthezia praelonga Douglas CTV Kirkpatrick 1953 b Phenacoccus madeiriensis Green CSSV Posnette 1950 Phenacoccus sp. CTV Kirkpatrick 1950 Pseudococcus sp. near comslocki CTV Kirkpatrick 1953 b Kuwana Puto barberi (Cockerell) CTV Kirkpatrick 1950 Steatococcus sp. CSSV Posnette 1951 Stictococcus sjostedti (Cockerell) CSSV Cotterell 1943 Bryocoropsis laticollis Sebum. CSSV Cotterell 1943 Sahlbergella singularis Haglund CNV Martini 1961 Cletomorpha lancigera F. CSSV Cotterell 1943 Selenothrips rubrocinctus Giard CSSV Cotterell 1943 Fulgeridae sp. CSSV Cotterell 1943 Gargara sp. CSSV Cotterell 1943 Criphyle sp. CSSV Cotterell 1943 Cercopidae sp. CSSV Cotterell 1943 Ricania mediana Mel. CSSV Cotterell 1943 Ricanopsis semihyalina Mel. CSSV Cotterell 1943 Ricanopsis nebulosa Mel. CSSV Cotterell 1943 Epitemna carbonaria Walk. CSSV Cotterell 1943 Pochazia fasciata F. CSSV Cotterell 1943 J) Evidence of transmission and non-transmission of CSSV available.