Claude Fauquet and Denis Fargette 1-3.” P< Laboratoire de Phytovirologie, ORSTOM, Abidjan, Ivory Coast

culture or together with maize, bananas can mosaic was first described (Musa acuminata Colla), peanuts (Ara- in 1894under the name Krauselkrankheit chis hypogaea L.), rice (Oryza sativa L.), and has since been reported through- beans (Phaseolus vulgaris L.), or other out Africa and in Madagascar, Zanzi- crops. . bar, Seychelles, India, and Java. In East Cassava is cultivated throughout the Africa, the disease was not reported to year between the latitudes 30’ north and cause serious losses until the 1920s. In south, up to a maximum altitude of West Africa, it was first recorded in the about 2,000 m. It tolerates drought but coastal areas of Nigeria, Sierra Leone, grows best where annual rainfall reaches and Ghana in 1929 and had spread 1,000-2,000 mm. Cassava grows well on northward by 1945. many soil types, with the exception of Responses to an international ques- hydromorphic soils, \Irhich are unsuit- tionnaire sent to all cassava-producing able. It is propagated vegetatively by African countries in 1987 (IO) showed stem cuttings, and the growth cycle gen- that ACMV is present wherever cassava erally ranges between 10 and 30 months. is grown on the continent (Table l), with Numerous constraints affect produc- the wettest regions the most seriously tivity of cassava in Africa. The most affected. The virus infects all cassava widely distributed pathogen is African cultivars, although susceptibility varies cassava (ACMV), dis- greatly. Incidence of ACMV infection is seminated by its vector, Bemisia often extremely high; a survey of 20 tabaci Genn., which is present in all cas- farms in Ghana revealed an average 96% sava-growing areas, and by man, in in- of plants infected. Similarly, the inci- center pf production (57 million tons fected cassava stem cuttings (Fig. 2). dence exceeded 80% in some districts of il were gqown on 7.5 million hectares in also infest cotton (Gossypium Kenya and was almost 100% in a few - 198Y). Gassava is produced for industrial hirsutum L.), tobacco (Nicotiana taba- small holdings. In Ivory Coast, the inci- - purpos{s in Brazil and as an export crop cum L.), and sweet potato (Ipomoea dence is nearly 100% in every cassava j 0- SZEZZ batatas (L.) Lam.), but cassava is the only plantation. Thus, ACMV is widespread U- m in Thailand, but in Africa it is grown --~ primarily for local food consumption. In cultivated host affected by ACMV. Other and destructive in the humid lowland 1 j-coc- IE-m-N nl fact, cassava is the most important food important diseases in Africa are vascu- areas of Africa. 3- lar bacteriosis caused by Xanthomonas g-o crop gròwn on the African continent, ex- 10-o ceeding yam (Dioscoreaspp.), 24 million campestris (Pammel) Dowson pv. mani- Symptomatology 1 ;= o tons; maize (zea mays L.), 12 million hotis (Berthet and Bondar) Dye; anthrac- The most visible symptom of the dis- tons; and pearl millet (Pennisetum glau- nose caused by Colletotrichuni gloeo- I IL___- ease is the expression of the characteristic 1 - cum (LI) R. Br.), 9 million tons. Pro- sporioides f. sp. nianihotis Henn.; and, leaf mosaic, and young plants are more duction per hectare averages 5-10 tons, to a lesser extent, cercosporiosis caused severely affected than old ones. Symp- but on /the basis of yields at research by Cercosporidium henningsii (Allesch.) toms range from barely perceptible mo- Deighton. The most significant pests are saic to stunting of the plant and extreme mealybug (Phenacoccus manihoti Mat.- reduction of the leaf blades (Fig. 3). The Ferr.), green spider mite (Mononychellus severity of symptoms is readily quanti- spp.), and the variegated grasshopper fied on a scale of 1-5, where O = no (Zonocerus variegatus L.). symptoms and 5 = leaves reduced .to veins (Fig. 4). Severity of symptoms Regional Importance of ACMV varies with the cultivar and increases with Present address of first author: Washington Uni- ACMV is considered to be of African plant age until about 60 days after versity, Department of Biology, Box 1137, St. Louis, origin and is unknown in South America. planting. Thereafter, symptoms are more MO 63130. Present address of second author: Scottish Crop Research Institute, Virology Division, However, similar symptoms are induced moderate or lessen or do not develop, DD2-5DA Dundee, Scotland. by the unrelated South American virus, depending on the cultivar, climatic con- cassava common mosaic virus, a potex- ditions, and season. Symptoms on fully ~ @ 1990The American Phytopathological Society virus having rod-shaped particles. Afri- expanded leaves do not change, however. No. 404 Plant Disease/Vol. 74 6 .".

iI

a

c I

.

1

Fig. 1. Principal routes of dissemination of cassava from South America to Africa and India, and distribution of cassava cultivation in Africa. The hatched lines mark the northern and southern limits of cassava production, and the dates are the years African cassava mosaic disease was first described in each country.

ent countries to range from 20 to 95% in 1969 was 29 t/ ha from those with mild Yield Losses (Table 1). The estimated average yield symptoms and only 9 t/ha from those ACMV is arguably the most important loss induced by ACMV is 50%. with severe symptoms. Differences in disease of virus etiology in Africa, but As intensity of symptoms increases, yield due to ACMV were greater than total losses are extremely difficult to yield of cassava declines dramatically those related to differences among culti- estimate. Yield losses with individual (Fig. 5). The mean yield of tubers from vars. Even cultivars considered to be re- cultivars have been reported from differ- a collection of cultivars in Ivory Coast sistant may suffer losses of 24-78% (1,6).

Fig. 2. African cassava mosaic virus is transmitted by (A) the adult whitefly (Bemisia fabaci) and (B) infected stem cuttings of cassava. Plant Disease/June 1990 405 With the moderately susceptible culti- euphorbiaceous species, Jatropha multi- the name Indian cassava mosaic virus var CB, yield losses range from O to 77%, fid L. Two other species, Hewittia sub- (ICMV) is now used for group C strains. depending on the mode and time of lobata (L.f.) Kuntze (Convolvulaceae) Mechanical transmission of ACMV infection (Fig. 6) and on environmental and Laportea aestuans (L.) Chew (Urti- was first reported in 1978. Purified prep- factors. Yield loss is greater (55-77%) in caceae), are suspected to be natural hosts arations from cassava plants in western cassava plants grown from infected for ACMV in Kenya and West Africa, Kenya contained a geminivirus, but Bock cuttings than in plants infected later by but the virus has not been transmitted et al (2) were unable to obtain this virus whiteflies (35-60%), even when infection from them back to cassava. from all infected plants. The etiology of occurs early. Infection that occurs by The geminate virus particles (Fig. 7) the disease therefore remained doubtful, means of vector 150 days or more after measure about 30 X 20 nm, and the coat and the geminivirus was named cassava planting has little effect on production protein has a molecular weight of about latent virus (CLV) (2) until it became (8). A virus-infected cassava plant remote 30,000. The particles contain one mol- apparent that the isolation problems from healthy cassava may produce 70% ecule of circular single-stranded DNA were due to differences among strains of less than its neighbors, but when all (Mr about 0.92 X lo6), and the genome the virus. CLV was shown to cause mo- plants in a plantation are infected, the consists of two circular molecules of saic and is correctly named ACMV (4). loss may be only 33% because the smaller similar size (15). In leaf tissue, the virus ACMV is mechanically transmissible infected plants compete less with one particles accumulate mainly in the nuclei from cassava to several solanaceous another than do healthy plants. of phloem parenchyma and of cortical plants, including species of Nicotiana On the basis of available data, the total and epidermal cells. and Datura, but successful transmission reduction of cassava yield in Africa in The type strain of ACMV (ACMV-T), back to cassava is difficult and is feasible plants derived from diseased cuttings is isolated in western Kenya, is serolog- only with very susceptible cassava at least 50%, or 50 million metric tons, ically related to all other whitefly-trans- cultivars. N. benthamiana Domin is the per year and may be equivalent to $2 mitted geminiviruses. ACMV isolates best source of virus particles for puri- billion (U.S.). from mosaic-affected cassava collected fication; chlorotic local lesions are frqm different parts of Africa and from followed by severe systemic leaf curling. India are partitioned into three groups D.stramonium L. can be used for local The Causal Agent on the basis of serology and DNA lesion tests; when it is inoculated with The viral etiology of African cassava hybridization (12). Group A includes some isolates, chlorotic and necrotic mosaic was first proposed by Storey in strains from Angola, Ivory Coast, Ni- local lesions are followed by systemic 1936 (17), who demonstrated in Tanzania geria, Congo, and western Kenya and veinbanding and leaf distortion. Virus that the disease was transmissible and also defective strains that do not produce infectivity in sap is unstable, being lost inferred that a virus was responsible. He virus particles; group B strains are from in a few days at room temperature or also studied virus transmission by B. coastal Kenya, Madagascar, and Ma- in 10 minutes at 55 C (3). tubaci (16,17). The establishment of lawi; and group C strains are from India transmission of ACMV in a persistent and Sri Lanka. The distribution of these The Whitefly Vector manner and the mode of transmission three groups may have resulted from the B. tubaci is a member of the Aley- were determined in 1958 and 1970, re- different routes of cassava dissemination rodidae, order Homoptera, and has four spectively. Storey showed that the white- (Fig. 1). As cassava was introduced into nymphal instars, of which only the first fly vector could introduce virus into the various countries that now constitute is mobile. The adult is about 2 mm long young leaves but not into mature leaves. its range, it may have become infected and difficult to identify; characters of the When plants were infected with a mild with different geminivirus variants that last nymphal instar are used in de- strain and then challenged with a severe were endemic in the three geographic terminations. The life span of B. tubaci strain, the mild strain did not always regions. Group C strains are not more is about 21 days and depends on cli- protect against the severe strain (18). closely related to other strains of ACMV matic factors, particularly temperature. ACMV, a geminivirus, infects seven than are different whitefly-transmitted Whether B. hancocki Corb also transmits Manihot species and a closely related geminiviruses found in other hosts, and ACMV has not been established. The adult whitefly needs about 3 hours to acquire the virus and, after a latent period of at least 8 hours, about 10 minutes to transmit the virus to healthy plants. Whiteflies are infective for about

Fig. 4. Rating of leaf symptoms of African cassava mosaic virus disease: O = no symptoms; 1 = faint mosaic; 2 = yellow mosaic, malformation, 5% size reduction; 3 = severe mosaic, distortion, reduced size; 4 =severe mosaic, severe distortion, Fig. 3. Symptoms of African cassava mosaic virus disease: (A) mild, (B) severe mosaic, up to 50% size reduction; and 5 = leaf (C) leaf curl, and (D)severely reduced leaf area. reduced to veins, 50-80% size reduction. 406 Plant DiseaseNol. 74 No. 6 7-9 days, and ACMV is not lost during and plant growth stage. Controlled flight the sources of inoculum-natural hosts molting. No transovarial transmission has been recorded only when wind speed or cassava. Epidemiological studies in occurs (5). The percentage of individuals drops below this limit, especially in early East Africa (1) and West Africa (6) ad- that become viruliferous when given morning, near ground level and within dressed these issues. access to infected cassava depends on the the cassava canopy. At greater wind Spatial distribution of ACMV. Studies cultivars tested and the whitefly popu- speeds, the insects are transported by the conducted in Ivory Coast have indicated lation but has been reported to range wind. that potential virus reservoir plants, such from 0.15 to 1.7% (6). Whiteflies are not distributed uni- as M. glaziovii Muell. Arg., around cas- Vector distribution, virus concentra- formly within cassava fields; their num- sava fields are unimportant epidemiol- tion, and leaf susceptibility to virus in- bers are highest on the upwind borders ogically and that cassava is the principal oculation are all related to leaf age. Up and lowest within fields, irrespective of reservoir for ACMV. J to 95% of adult whiteflies found on field size or whitefly population. The size Within newly infected cassava fields cassava are concentrated on the abaxial of vector populations is positively cor- in Ivory Coast, the distribution of the surface of the five youngest leaves of each related with virus spread about 1 month disease shows a dispersion gradient of shoot. Despite symptoms, virus particles after invasion, which corresponds ap- ACMV correlated with the vector dis- cannot be detected in leaves older than proximatively to the time necessary for tribution. Incidence of the disease is the seventh from the apex (Fig. S), and symptom development. The environ- higher along the upwind edges than on only the five youngest leaves of each mental factor that correlates best with the downwind edges of the field. This shoot are susceptible to virus inocu- fluctuations in whitefly population is distribution shows up as a curvilinear lation. temperature, also with a time lag of 1 gradient of infection in the direction of Spread of the disease in time and space month, the approximate generation time the prevailing wind (Fig. 10). Such a is related to the movements of adult of the insect. gradient occurred in all fields inves- whiteflies. The flight speed of B. tubaci tigated despite greatly differing areas and has been calculated to be about 0.2 m/ exposure conditions. Plant density also sec, but the insects can control their flight Epidemiology influences field contamination, i.e., infec- only in air with reduced wind speed, such For effective control of ACMV, one tion progresses fastest with low density as occurs within the plant canopy must establish whether man or the and incidence of disease is lowest with (Fig. 9). The depth of this comparatively whitefly is more important in dissemi- high density. motionless layer depends on wind speed nating the virus and also must identify Primary spread of ACMV by viru- liferous immigrant insects landing in healthy cassava fields has been distin- guished from secondary spread from diseased plants within the fields. Primary spread accounts for about 70% of vector transmissions. Within a healthy cassava field, the dispersal gradient from a source of contaminated plants, although occur- ring in all directions, does not exceed several meters and probably is related to the relatively limited flight of white- flies within fields. Cassava is often cultivated with other + +.- r-; food crops, and experiments in Ivory + Coast with the association of cassava and +s maize show a great influence of maize I on the spatial distribution of ACMV (7). .m. .I..-.. .I. .s. d .5 1 1.5 i 2.5 i 4.5 More detailed studies are necessary to 3.5 4 6 evaluate precisely this influence on the INDEX OF SYMPTOM SEVERITY spread of the virus. Fig. 5. Relationship between yield and symptom severity in 49 cassava cultivars in Temporal spread of ACMV. Temporal Ivory Coast. (Courtesy R. Vandevenne) spread of ACMV depends on numerous factors, some of which are linked or interact, as illustrated in plantings of the

0 60 90 120 150 180 210 240 270 300 330 DAYS AFTER PLANTING Fig. 7. Electron micrograph of particles Fig. 6. Relationship between yield of cassava cultivar CB and time and mode of infection of African cassava mosaic virus. Scale bar with African cassava mosaic virus. = 100 nm. Plant Disease/June 1990 407 cultivar CB in Ivory Coast during 1983-1986. Spread of disease varied ITransmission efficiency I greatly from month to month and 4 71 0 12.3 O481216 O 0.2 0.4 0.6 0.8 showed an annual periodicity with sea- sonal variations (Fig. 11). Up to 87% of the cassava planted in March was in- fected within 2 months, but incidence in August plantings did not exceed 4%. The susceptibility of plants decreases with age, and little infection occurs after 3 months (Fig. 12). Disease incidence largely reflects fluctuations in whitefly populations but also depends partly on variations in climatic factors, including temperature, rainfall, and wind. Rela- tionships among disease incidence, vec- tor populations, plant growth, and cli- matic factors are complex. Temperature is positively related to vector populations Fig. 8. Susceptibility of cassava leaves of increasing age (F1 to F11) to inoculation and disease incidence. by insects determined by (left) the number of infected plants per 10 inoculated plants Regional spread of ACMV. Extensive per leaf, (center) the mean number of adult whiteflies per leaf, and (right) the virus tests of cultivars and locations were content, expressed as average ELISA absorbance values at 405 nm. conducted from 1977 to 1984 in East Africa on the coast of Kenya. There was little spread of ACMV (1-2%) into plots of initially mosaic-free selected cassava OF NUMBER OF WHITEFLIES FLYING NUMBER WHITEFLIES FLYING hybrids, irrespective of plot size, loca- IN A CASSAVA FIELD AT EACH HEIGHT IN EACH DIRECTION I Il I tion, and annual or regional climatic variations. Some local cultivars seemed 10 6 to become infected no more often than 6 hybrids selected for ACMV resistance. 4 In Kenya, therefore, infection prevails because farmers do not discrimate be- 2 o tween contaminated and healthy plants S SE E NE N NW W SW when choosing cuttings. There was little -SOUTH-WEST PREVAILING WIND - spread of ACMV to highly susceptible 1W pq cassava cultivars at sites isolated from 801 areas of dense cassava cultivation, even 60. though the prevailing wind traversed 40. many small plots of diseased cassava 20. plants. Spread of ACMV for consider- O, S SE E NE N NW W SW able distances appeared to be limited. In io ...... contrast, short-range dispersal was 0.2 2.0 b 50 100 150 200 250 300 WINO SPEED mls NUMBER OF WHITEFLIES WIND DIRECTIONS apparent at several sites. At the same site, infection rates for the same susceptible cultivar ranged from 8 to 70%. The plot Fig. 9. Movement and spread of whiteflies in a cassava field as related to wind speed, with lowest incidence of ACMV was wind direction, and cassava canopy and estimated by the number of trapped whiteflies in each direction at different levels, under and above the canopy. isolated, whereas the plot with the high- est incidence was surrounded by fields of diseased cassava. The results confirm that ACMV can be succesfully eradicated in Kenya by using healthy cuttings of ioo- LWIND DIRECTION -35 local or selected cultivars that have some 90 '- resistance to ACMV. '-30 EO - In West Africa, trials conducted in many locations in Ivory Coast revealed 0 70- -25 5 sites with consistently high (82%) or low - =i (1%) incidences of ACMV, indicating $- $- 60'- -20 ; W -m regions of high and low inoculum avail- 5 50'- ability. Although the number of experi- o -15 u 40.- : mental locations was restricted, it seemed that the inoculum pressure at any one z '- 30 -10 --I5 site may have been related to the overall I 20 '- density of cassava cultivation in the re- .-5 gion. At the same site, incidences of IO '- ACMV were high (84%) in susceptible I cultivars and low (10%) in resistant O I Io cultivars, and some of the local cultivars 5 15 25 35 45 55 65 75 of East or West Africa origin appeared DISTANCE FROM EDGE OF FIELD (m) Fig. 10. Incidence of African cassava mosaic virus infection and number of whiteflies to be as resistant as selected cultivars per plant 3 months after planting, along the axis of the prevailing wind direction in introduced from Kenya or Nigeria. Some a cassava field. short-range dispersal of ACMV was

408 Plant DiseaseNol. 74 No. 6 ~~

demonstrated, and the influence of in- sistant cultivars are used or disease produce edible roots (13). This breeding fected cassava surrounding virus-free incidence is low, but that spread by program resulted in several cultivars plots was evaluated in the vicinity of few whiteflies is also important. Cassava considered by breeders to be “very highly or many fields of infected cassava in the serves as a reservoir for both virus and resistant.” Seeds from these cultivars sites with high or low incidences of vector in East and West Africa. were used at IITA as the source of ACMV. The transmission ability of the resistance to ACMV. The IITA program whiteflies at each site was calculated as (1 1) incorporates ACMV resistance and the ratio between the number of white- Resistance and Selection several other factors, especially yield. flies per cassava plant and the number In East Africa in 1938, Storey and Several of these cultivars are now avail- of infected plants per field. This ratio, Nichols (18) began the first cassava able to farmers; the most resistant to 240 days after planting, was in the range selection program against ACMV, and ACMV are TMS 30572, TMS 30395, 3 of 40-80 for sites with diseased cassava this program continued until 1957 TMS 30001, and, most recently, TMS upwind and 300-1,000 for sites with no (13,14). Seeds from the program were 4(2)1425 (Fig. 13). Initially, East African infected plants within a few kilometers distributed for use in many African coun- cassava breeders selected for field upwind, indicating that whiteflies iso- tries and provided the main source of resistance according to how long plants lated from infected cassava plants are less resistance used in the breeding program remained healthy after planting; later viruliferous. at the International Institute for Tropical they selected according to symptom in- The epidemiological studies conducted Agriculture (IITA), Ibadan, Nigeria, in tensity. In West Africa, the criteria for in East and West Africa led to the same 1971. selection were based on intensity of conclusion, i.e., that major outbreaks of Initially, the Javanese cultivar F279 symptom expression, ranked according ACMV within a field are a consequence and several African cultivars considered to a scale resembling the Cours scale of primary spread from outside sources. to have ”some resistance” to ACMV were (Fig. 4). Resistance to ACMV seems to Sanitation is especially important when- used to create the hybrid 37244E (Fig. be polygenic and recessive and is ever planting material is resistant to 13). The level of resistance obtained was positively correlated with resistance to ACMV, but diseased plants upwind are inadequate, however, and interspecific bacterial blight (1 1). influential in the contamination of virus- crosses were then made between M. escu- In Ivory Coast, 54 cassava cultivars free fields. These studies show that cul- lenta and M. glaziovii or M. melanobasis have been tested to determine the differ- tural practices are significant in dis- Muell. Arg. Each cross was followed by ent components of ACMV resistance (9). seminating ACMV, especially when re- three backcrosses to M. esculenta to The cultivars originated in nine locations on three continents and included local and selected cultivars as well as intra- I”” I specific and interspecific hybrids between 90 M. esculenta X M. glaziovii and M. escu- X 80 lenta M. melanobasis. Six different h components of resistance were consid- v8 70 ered: field resistance (percentage of in- 60 fected plants), vector resistance (number

50 of adult whiteflies), inoculation resist- 5 ance, virus resistance (virus content esti- 5z 40 mated by ELISA), symptom intensity, 30 and virus diffusion resistance (develop- ment of symptoms with time). Little 20 correlation was found among field re- 10 sistance, vector resistance, and virus diffusion resistance, but field resistance O AJAODFAJAODFAJAODFA was highly correlated with the other three components, particularly symptom in- 1983 1984 1985 1986 Fig. 11. Incidence of African cassava mosaic virus, expressed as frequency of infection tensity. Multicomponent statistical anal- 2 months after planting, for successive monthly plantings from 1983 to 1986. yses were used to obtain precise descrip- tions of the different cultivars and com- ponents of resistance, and hierarchical classifications were assigned to different groups. The most field-resistant group includes all hybrids from East and West Africa, local cultivars from Kenya and India, and the outstanding Aipin valenca chosen by Storey as the parent for many crosses. Transmission of ACMV from one generation to the next through cut- tings has not been thoroughly studied for cultivars, but several experiments indi- cate that failure to persist is a trait that might be very important. A study of genetic ‘relationships be- tween M. esculenta and M. glaziovii in Ivory Coast showed that these species hybridize naturally and that intermediate forms occur in Africa. Biochemical 1 2 3 4 5 6 markers are similar in both local and AGE OF THE PLANTS IN MONTHS selected ACMV-resistant cultivars and Fig. 12. Decrease of susceptibility to African cassava mosaic virus in cassava plants are correlated with at least one crossing 1-6 months old. with M. glaziovii. Plant Disease/June 1990 409 Control Strategies Two control strategies apply: 1) the tion of sanitation methods are obvious Sanitation and breeding for resistance, use of cultural techniques such as and considerable. Virus-free cassava cut- the primary practices for control, have sanitation, which includes planting tings take root and grow faster, produce been investigated since 1936. Storey ob- healthy cuttings, and 2) the use of re- more tubers, and result in a higher yield, served that ACMV was less frequent in sistant cultivars to supersede susceptible even if they are inoculated later by highland areas of Tanzania than on the ones. Combining these two strategies whiteflies. The effectiveness of this coast and stated that “it is perfectly undoubtedly would result in consider- method depends on the rate of recon- feasible . . . to establish healthy plots able improvement of cassava production tamination of virus-free planting mate- and to maintain them, by inspection and in Africa, especially since ACMV causes rial, the availability of stocks of cuttings roguing, practically disease-free.’’ When debilitation and predisposes to other for desired cultivars, and the ease with Bock (1) conducted a series of epide- pests and diseases of cassava. which the cuttings can be produced. miological trials in Kenya during 1974- Cultural techniques. With some cul- Because cassava is grown mostly by small 1981, sanitation in resistant cultivars tivars, healthy cuttings can be selected farmers, large-scale sanitation can be combined with isolation from the in- from symptomless plants in fields where truly effective only if undertaken and fected fields controlled ACMV. infection prevails. Such careful selection, promoted by the farmers themselves. There has been little attempt in many called “reversion” in Africa, has enabled Among local and selected cultivars, parts of Africa to control ACMV. Cas- large-scale production of healthy cassava plants that show reversion can be sources sava mosaic is so widespread that most p1,ants of several different cultivars in of selected material because reinfection producers are indifferent to the disease Kenya and Ivory Coast. With cultivars is slow. Selection of healthy plants relies and are not aware of the considerable for which no reversion has been ob- only on recognizing the characteristic yield losses sustained. Farmers grow gen- served, the techniques of meristem cul- symptoms, but application requires etically very diverse mixtures of cultivars, ture and thermotherapy must be used. adoption of strict cultural methods by often including some that would enable For all cultivars, a large-scale program local farmers in the choice of symp- them to control the disease effectively. for the sanitation of cassava by vegetative tomless plants. An effective program is Simple techniques such as selection of micropropagation in vitro might be possible only through development and healthy cuttings for replanting could initiated. implementation with extension services and personnel. rapidly reduce the impact of ACMV. The potential benefits of the applica- Some East African countries, includ- ing Kenya, Uganda, Tanzania, and 4 E Manihot esculenta X Manihot esculenta -b 34 2 4 Malawi, could develop action programs AFRICAN CULTIVARS F279 JAVA based simply on roguing and sanitation because such extension services and x 3 backcrosses to personnel are available. The research a Manihot esculenta Manihot melanobasis -+ 5 5 4 3I O 5@ work begun in 1935 has produced cul- AFRICAN CULTIVARS Manihot esculenta tivars sufficiently resistant that recon- !a tamination will not be a serious problem. ‘ Manihot esculenta X Manihot glaziovii Moreover, whiteflies multiply slowly or AFRICAN CULTIVARS not at all at high altitude, thus making a sanitation easier. I 3 backcrosses to Finally, if cassava is the main reservoir Manihot esculenta of virus and vector, inoculum will in- crease with the intensity of cropping. As 1 more diseased cassava plants are grown, sanitation techniques will be more dif- 461 @@/ZU “Modemtely msistan!“ ficult to apply. These factors explain the 4 high incidence rates and difficulties 5 31 8 13 “Vevery highly resistant“ applying sanitation methods in West Africa, where resistant cultivars were not SEEDS developed until 1975-1980. The dry sea- IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII son is relatively short and frequently in- terrupted by rains that favor multipli- cation of the vector, high-altitude culture Manihot esculenta1 X Manlhot esculenta is not possible, and cultivation of cassava is very intense. Roguing of diseased 4 plants, selection of planting date, plant density, and disposition and size of 1 plantings all have been effective under ”Modemtely resistan!” 58 30 8 X Manihot escdenta some conditions, but they require levels zb NIGERIAN CULTIVARS of technical input not available to local farmers and possible only in industrial a plantations. Nevertheless, these control x Manihot esculenta techniques could become effective under SOUTH AMERICAN CULTIVARS certain conditions and merit greater con- sideration. 1 Resistant cultivars. Many resistant c “Highlyresistan! UMS 30542 cultivars are available, either produced and high yield” UMS by breeders or selected in natural con- 30395 ditions. Dissemination of such cultivars UMS 30001 UMS currently is restricted by insufficient 4(2)1425 knowledge and scanty distribution of Fig. 13. Breeding and selection schemes for resistance to African cassava mosaic v%Ïs resistant plant material, although efforts in East and West Africa. are being made to improve the situation. 410 Plant DiseaseNol. 74 No. 6 The use of resistance is affected also by and the possibility for control of ACMV attitudes of local farmers who demand, is slight. Only combined research and in addition to disease resistance, suitable development can improve control of root processing qualities and, above all, ACMV in the traditional environment in organoleptic qualities. These criteria are which almost all African cassava is extremely variable and depend on prefer- produced. ences of ethnic groups. Cultivars bred at research stations often do not meet Literature Cited local needs and desires and therefore 1. Bock, K. R. 1983. Epidemiology of have not been accepted. For example, ca5sava mosaic disease in Kenya. Pages resistant cultivars from IITA, including 337-347 in: Plant Virus Epidemiology. R. TMS 30572, have been distributed T. Plumb and J. M. Thresh, eds. Black- primarily in Nigeria, in part for technical well Scientific Publications, Oxford. reasons but also because the cultivars 2. Bock, K. R., Guthrie, E. J., and Meredith, have characteristics favoring gari (local G. C. 1978. Distribution, host range, processed cassava) production, which is properties and purification of cassava much appreciated in Nigeria. However, latent virus, a geminivirus. Ann. Appl. Claude Fauquet Biol. 90:361-367. Dr. Fauquet has been a plant virologist the cultivars lack the processing qualities 3. Bock, K. R., and Harrison, B. D. 1985. with ORSTOM (Institut Français de required in Benin and Togo. Thus, the African cassava mosaic virus. No. 297 Recherche Scientifique pour le DtSvel- use of resistant cultivars as a strategy of in: Descriptions of Plant Viruses. oppement en CooptSration) since 1972. control will become widespread only Commonw. Mycol. Inst., Kew, Surrey, He was awarded a Ph.D. degree in when breeders take into account proc- England. 4 pp. 1974, a These d’Etat degree in 1985, essing criteria and local preferences, 4. Bock, K. R., and Woods, R. D. 1983. and the Diplome d’Habilitation B promotion through the dissemination of Etiology of African cassava mosaic Diriger des Recherces (H.D.R.) in 1987 information, and acceptance of im- disease. Plant Dis. 67:994-995. from the Louis Pasteur University of proved local cultivars. Techniques of bio- 5. Dubern, J. 1979. Quelques propriétés de Strasbourg. Since 1974, he has been la mosaique Africaine du manioc. I. working in Ivory Coast, West Africa, technology may enable insertion of re- Transmission. Phytopathol. 2. 96:25-39. taking inventory of viral diseases of sistance genes into traditional cultivars 6. Fargette, D. 1985. Epidemiologie de la cultivated crops, classifying plant that are susceptible to ACMV but suit- mosaique Africaine du manioc en Ivory viruses by means of the amino-acid able in other respects. Coast. Ph.D. thesis. Science and Tech- composition of coat proteins, and nology University, Montpellier. 203 pp. studying the epidemiology of African Control strategies should be used in 7. Fargette, D., and Fauquet, C. 1988. A cassava mosaic virus. Since 1988, he concert because no cultivar is immune preliminary study onfthe influence of has been studying at Washington Uni- to ACMV and even resistant cultivars intercropping maize and cassava on the versity in St Louis and developing the may become infected. When infection spread of African cassava mosaic virus Cassava-Trans Project, using biotech- occurs, yield losses can be considerable, by whiteflies. Aspects Appl. Biol. 17:195- nology techniques and the “coat pro- even with resistant cultivars. In East 202. tein strategy” to control African Africa, resistance and sanitation have 8. Fargette, D., Fauquet, C., and Thouvenel, cassava mosaic virus and cassava been used in combination to control J-C. 1987. Cassava crop losses due to the common mosaic virus. He has been a ACMV, and experiments in Ivory Coast African cassava mosaic virus. Trop. Pest member of the Committee of Virus Manage. 3497-99. Epidemiology of ISPP since 1983 and have shown that virus-free cassava can 9. Fauquet, C., and Fargette, D. 1986. A secretary of ICTV since 1987. be grown year-round using a cultivar summary of the epidemiology of African such as Aïpin valenca. Control of ACMV cassava mosaic virus. Pages 1-30 in: Proc. depends on the amount and efficacy of Plant Virus Epidemiol. Workshop 3rd. inoculum, the resistance of cultivars, and 10. Fauquet, C., and Fargette, D. 1987. the agronomic techniques employed. African Cassava Mosaic Disease. Inter- Effectiveness of inoculum is a function national Investigation in Africa. CTA of biological factors, such as intensity of Eds., Ede, Netherlands. 117 pp. cassava cultivation and density of vector 11. Hahn, S. K. 1980. Breeding cassava for resistance to cassava mosaic disease. population, and of climatic factors and Euphytica 29:673-683. geographic location, especially altitude. 12. Harrison, B. D., Lennon, A. M., Thus, the choice of cultivar and the cul- Massalski P. R., Robinson, D. J., and tural techniques to be used will depend Thomas, J. E. 1986. Geographical vari- on expected effectiveness of inoculum. ation in African cassava mosaic virus. The types of resistance are important; Pages 9-11 in: Proc. Plant Virus for instance, the cultivars Aipin valenca Epidemiol. Workshop 10th. and TMS 4(2)1425 are resistant to spread 13. Jennings, D. L. 1957. Further studies in due to reversion, which effectively results breeding cassava for virus resistance. East Afr. Agric. J. 22213-219. in a “genetic cleanup” at each new gen- 14. Nichols, R. F. W. 1947. Breeding cassava Denis Fargette eration, even if the source plant is in- Dr. Fargette has been a plant virologist for virus resistance. East Afr. Agric. J. fected. This aspect of control is being 15:154-160. with ORSTOM since 1980. He received pursued in research studies. Finally, the his Ph.D. degree from the Science and 15. Stanley, J., and Gay, M. R. 1983. Technology University of Montpellier cultural component depends mostly on Nucleotide sequence of cassava latent in 1985. In Ivory Coast, his research human and developmental factors and virus DNA. Nature 301:260-262. activities mainly concerned epidem- on international cooperation. In any 16. Storey, H. H. 1934. Report of the plant iology of African cassava mosaic virus, given cassava-producing area, different pathologist. East Afr. Agric. Res. Stn. vector transmission, and resistance sets of conditions apply to industrial cas- Amani Annu. Rep. 6:l-10. mechanisms. He has also been in- 17. Storey, H. H. 1936. Virus diseases of East sava plantations, where many cultural African plants. VI. A progress report on terested in classification of plant techniques and cultivars are options and viruses and is currently working at the studies of the diseases of cassava. East the chances for controlling ACMV are Afr. Agric. J. 2:34-39. Scottish Crop Research Institute in good, and to plantations of local farmers, Dundee, Scotland, on characterization 18. Storey, H. H., andNichols, R. F. W. 1938. of geminiviruses. where techniques other than sanitation Studies on the mosaic of cassava. Ann. are few, the choice of cultivars is limited, Appl. Biol. 25:790-806. Plant Disease/June 1990 411 .. i

..-