COFFEE BERRY DISEASE IN KEf'YA

CENTRALE LANDBOUWCATALOGUS

OOOO 0086 7297 Promotor:dr.ir .J . Dekker,hoogleraa r ind efytopathologi e aand e Landbouwhogeschool hn <9zoi , ys3>

H. VERMEULEN

COFFEE BERRY DISEASE IN KENYA

Proefschrift terverkrijgin gva nd egraa dva n doctori nd elandbouwwetenschappen , opgeza gva nd erecto rmagnificus , dr.H.C .va nde rPlas ,hoogleraa ri nd eorganisch e scheikunde, inhe topenbaa rt everdedige n opwoensda g2 8februar i 1979de snamiddag st evie ruu r ind eaul ava nd eLandbouwhogeschoo lt eWageninge n

j£>; 'op rvt L- :'•'•• •;;;:E K

LANDBt '^'Vit!>..SCHOO L Ter nagedachtenisva nmij nouder s

BIBLIOTHEEKLH . 3 1 JAN. 1979 ONTV.TIJOSCHR .AOM .

Thou,wh oca nmak emountain smov e andriver sflow , acceptou rsacrific e andgiv eu srain .

Excerpt from a Kikuyu-prayer for rain. nvzcri 7^3

STELLINGEN

1. Zolangnie tovertuigen di svastgesteld ,da te ree nrelati ebestaa ttusse n Glomerella aingulata en Colletotrichum coffeanum inKenia ,moe tworde naangenome n datd eascospore nva n G. aingulata inda tlan dgee nro lspele ni nd eepidemiologi c vand ekoffiebesziekte .

Dit proefschrift

2. De Colie totriefewm-populatie ind ebas tva nkoffiebome ni sonde rinvloe dva nlang - durigee nintensiev efungicidenbespuitinge ni nKeni averschove ni nd erichtin gva n depathogen ecomponent , C. coffeanum, deverwekke rva nd ekoffiebesziekte .

Dit proefschrift

3. Laboratoriumbepalingenva nd etotal econidienprodukti eva nd ebast-populatie sva n Colletotrichum-soorten zeggenweini gomtren td et everwachte noogstverlieze nte n gevolgeva nd ekoffiebesziekte ,noc hva nd eeffectivitei tva neerde ruitgevoerd e bespuitingenme tfungiciden .

Dit proefschrift

4. Opd etakke nva nd ekoffieboo mi she tvoorkome nva n C. coffeanum beperktto tdi e bastgedeeltenwaa rd evormin gva nsecundair ekurklage ni nd ecorte xno gne tnie t ofjuis ti sbegonnen .

Ditproefschrif t

5. Inverban dme trecent ewaarneminge nove rhe toptrede nva nflagellate nva nhe tge - slacht Phytomonas inkokos -e noliepal mi she tgewens td ereed si n193 1doo r Stahelontdekt efytopathogen eflagellate noo ki nNederlan daa nee nhernieuw don ­ derzoe kt eonderwerpen .

Parthasarathy,M.V. , Slobbe,W.G . van & Sondant, C. Science 192 (1976). Slobbe,W.G . van,Parthasarathy ,M.V . & Hesen,J.A.J . Principes,22 , 1 (1978). Stahel, G. Phytopath. Z. 1 (1931), 5 (1932), A (1933). Waters,H . Ann. appl. Biol. 90 (1978). 6. Dealgemee ngeaccepteerd eopvatting ,da t Trypanosoma gambiense nietmee rvoorkcm t alsverwekke rva nslaapziekt ei nd eRoovalle ibi jhe tVictoriamee ri nOos tAfrika , iswaarschijnlij knie tvoldoend egefundeer dgezie nd eparasitologisch ee nbio - chemischeeigenschappe nva nd etrypanosomen ,geisoleer dui tgeinfecteerd etse-tse - vliegenui tdi tgebied .

Allsop,R .& Baldry ,D.A.T .Bull .Wld .Hlth . Org.4 7 (1972). Gibson,W . Trans. Roy. Soc. Trop.Med . &Hyg . (inpress ) (1979). Goedbloed, E.,Ligthart ,G.A . &Minter , D.M. Trans. Roy. Soc. Trop.Med . &Hyg . 65 (1971).

7. Eenzelfstandig eontwikkelin gva nee npootaardappelprodukti ei nIndonesi eza l alleenmogelij kzij nindie ni svastgestel dho ehe toptrede nva ninsecten ,di ed e verschillendeaardappelvirusse noverdragen ,o peffectiev ee neconcmisch ewijz eka n wordentegengegaan ,waarbi jvoora lhe tgedra gva ndez evectore ni naanmerkin gmoe t wordengenomen .

Dedoo rPostgat egeuit everonderstelling ,da tvrijlevend estikstofbindend e bacterienwaarschijnlij kslecht si nrelati eme td eplanteworte lhu nfuncti ekunne n vervullen,gaa tgehee lvoorbi jaa nhu nbetekeni si nd efyllosfeer .

Bessems,E.P.M . Nitrogen fixation in the phyllosphere of Gramineae. Agric. Res.Rep . 786 (1973). Postgate,J.R . Biological nitrogen fixation. In: Companion tomicrobiology , selected topics for further study (EdsA.T . Bull & P.M. Meadow)Longman ,Londo n- Ne w York (1978). Ruinen,J . Nitrogen fixation by free-living micro-organisms (Ed.W.D.P . Stewart). Inter­ national Biological Programme Vol. 6 (1975).

9. Voorhe tinstandhoude ne naccumulere nva nvoldoend ekermi se nervarin go phe tge ­ biedva nlandbou wi nd eDerd eWereld-lande ni see ngerich tbelei dnoodzakelijk . Hetontbreke nva nzo' nbelei di nNederlan ddoe tafbreu kaa nd ekwalitei tva nd e agrarischeontwikkelingshulp . 10. Deprojectduu rvoo r toegepast landbouwkundig onderzoek indoo rNederlan dgefinan - cierde ontwikkelingshulpprojecteni nd eDerd eWerel d isveela l tekort .He tbelan g end emoeilijkheidsgraa dva ndergelij k onderzoekword t doord ebeleidsfunctionaris - seni nNederlan d onvoldoende onderkend,hetgee nnadeli g isvoo rhe trendemen t van deontwikkelingshulp .

11. Lienbelangrij kdee l vand eNederlands e literatuur overbiologisc h enagrarisc hon ­ derzoekverrich t inIndonesi e isva nblijvend ebetekeni s voorda t land.Doo rd e nog aanwezigekermi sva nd eNederlands e taalbi j de oudere Indonesischevak - specialisten kandez e literatuurn uno gme t geringe financielemiddele nworden.ver - taald inhe t Indonesisch.

12. Moderne Nederlandse schrijversverdiepe n zichveelvuldi g ineige njeugdtrauma' se n -ervaringen.He t isechte r tevreze nda thieraa nonvoldoend e inspiratie kanworde n ontleend vooree nblijven d schrijversschap.

13. Deopvattinge n ind eNederlands e samenlevingme tbetrekkin g tothe t kolonialisme hebbennie t geleid totee nafwijze nva nmodern evorme nervan ,zoal sdi tonde rmee r totuitin g komt ind e sociaal-maatschappelijkewaarderin gva nd egastarbeider .

14. Het toelatento t tentoonstellingen enhe tverstrekke nva nprijze naa nhonde nme t gecoupeerde oreno pNederlands e kynologische exposities is instrij dme the t sinds 1961 geldende verbod inNederlan d ophe tcoupere nva noren .

H.Vermeulen , "CoffeeBerr yDiseas e inKenya" , Wageningen, 28februar i1979 . CONTENTS

PREFACE

1. INTRODUCTION 1

2. COFFEE CULTURE INKENY A 3

2.1 Biologyo f Coffea spp. 3 2.1.1 Taxonomy 2.1.2 Varieties and cultivars 2.1.3 Morphology 2.1.3.1 General habitat 2.1.3.2 Flowering 2.1.3.3 Fruit development stages 2.2 Arabica coffeei nKeny a 6 2.2.1 Introduction, spread and importance of arabica coffee in Kenya 2.2.2 Environmental and cultural aspects of coffee growing in Kenya 2.2.2.1 Environment a.Climat e b. Soils 2.2.2.2 Aspects of coffee cultivation a.Prunin g b.Plantin g distance c.Mulc han dfertilizer s d.Wee d control e. Irrigation £.Shad e trees 2.2.3 Organization of coffee research 2.2.4 General disease and pest situation in Kenya, 2.3 Terminology 13

3.LITERATUR E REVIEWO FCOFFE E BERRY DISEASE 16

3.1 Discovery andsprea d ofcoffe e berry disease inKeny a 16 3.2 Economic losses inKeny a 16 3.3 Symptomso fcoffe e berry disease 17 3.4 Occurrenceo fcoffe e berry disease outside Kenya 17 3.5 Colletotriahum spp.o ncoffe e 18 3.6 Taxonomy of Colletotrichum spp. 19 3.7 Other host plantso fcoffe e berry disease 20 3.8CB Dresearc h inKeny a 21 3.8.1 Research between 1922 and 1950 3.8.2 Research between 1950 and 1966 3.8.3 Research after 1966

4.MATERIAL SAN DMETHOD S 28

4.1 Coffee cultivars 28 4.2 Fungi 28 4.3 Fungicides 28 4.4 Isolation offung i 30 4.5 Inoculation methods 30 4.6 Assessmento fconidia l germination andpathogenicit y 30 4.7 Production ofconidi a onth ebar k andthei r pathogenicity 32 4.8 Assessmento fCB Dlevel s infiel d experiments 34 4.9 Screening offungicide s 35

5.ANATOMICA L ANDMYC0L0GICA L ASPECTS 36

5.1 Coffee berry diseasei nKenya .I . Colletotrichum spp.colonizin gth e barko f Coffea arabica. (reprinted from Neth. J. Plant Path. 76 (1970) 277-284). 36 5.2 Coffee berry disease inKenya . II. Therol eo f Glomerella cingulata inth e Colletotrichum population,colonizin g thebar ko f Coffea arabica. (reprinted from Neth. J. Plant Path. 76 (1970) 285-292. 44 5.3 Ascospore Production,Discharg ean dInfectio n by Glomerella cingulata causing Coffee Berry Disease. (reprinted from Nature, Vol. 214, No. 5093, pp 1144-1145, June 1967) 52 5.4 Conidial germination andinfectio n 55 5.4.1 Introduction 5.4.2 Results of germination and infection experiments 5.5 Discussion 59

6. SP0RULATI0N ANDPROBABILIT Y OFINFECTIO N INRELATIO N TORAINFAL LAN D TIMINGO FFUNGICID E APPLICATIONS 64

6.1 Introduction 64 6.2A shor t noteo nsom e coffee spray timing trials against coffee berry disease. (reprinted from Kenya Coffee, 31 (1966), 368: 343-345) 66 6.3Sporulatin g capacity inrelatio n torainfal l andfungicid e applications 72 6.4Pathogenicit y ofconidi a produced onbranche s inrelatio n tosporula ­ ting capacity 73 6.5CB Dincidenc e inth efiel do flabelle d branches 75 6.6 Discussion 71

7. FUNGICIDE SCREENING TRIALS 80 7.1 Screening fungicides for control ofcoffe e berry diseasei nKeny a (reprinted from Expl. Agrio. 4 (1968), 255-261) 80 7.2 Discussion 87

8. EFFECTSO FCULTURA L PRACTICESO NTH EINCIDENC EO FCOFFE E BERRY DISEASEO N TREESAN D THE SUSCEPTIBILITYO FBERRIE S 89 8.1 Introduction 89 8.2 Results 89 8.3 Discussion 91

9. GENERAL DISCUSSION AND CONCLUSIONS 93

10. SAMENVATTING 99

11. SUMMARY 103 12. BIBLIOGRAPHY 104 PREFACE

Althoughman yyear shav epassed ,I a mstil lgratefu l toth estaf fo f theCoffe e ResearchStatio na tRuir ufo rthei rhel p inth eperio d 1964-1966 .Furthermor eI ammos t thankful toth eSenio rPlan tPathologis to fth eNationa lAgricultura l Laboratories, Mr J.J. OndiekiM.Sc .an dhi s stafffo rthei rhel pan dhospitalit y inth eyear s 1967-1969 .Th efiel dobservation s andcollectio no fcoffe emateria l inEmb uan dMer uwer emad epossibl eb y thecooperatio no fth e fieldofficer so f theMinistr yo fAgricultur e inthos eareas .Finall y Iwis ht oacknowledg eth eop ­ portunity thatth eGovernmen to fth eRepubli co fKeny aan dth eKingdo mo fth e Netherlandsgav em e forthi s research incooperatio nwit h thethe nEas tAfrica n Agriculturean dForestr y ResearchOrganizatio nunde r thesupervisio no fth e InternationalAgricultura lCentre ,Wageningen ,th eNetherlands . Betweenm y departurefro mKeny a inSeptembe r 1969an dth epublicatio no f thismanuscrip t in 1979,m y secondmentt o Indonesia from 1970-1977 delayedpre ­ parationo fm yCB Dresearc hfinding s in toto. Afterm y returnt oth eNetherlands , Iwa sposte d toth eLaborator y ofPhytopathology ,Agricultura lUniversity , Wageningen,m y customary refugedurin g leavesfro moverseas .Professo rD rI r J. Dekkeran dhi sstaf freceive dm e againmos tkindly ,providin g all facilities andhelp . The ideao fpreparin g aful laccoun to fm y research inKeny awa s reborni n theearl ypar to f 1978an dProfesso rDekke rsupporte d itwholeheartedly . Ia m mostoblige dfo rth eencouragemen to fm y oldteacher ,Professo rD rA.J.P .Oor t andfo rth ehel p Ireceive d fromD rJakob aRuine nan d IrJ.H .va nEmden .Th econ ­ tinued interest andadvic eo fProfesso rDekke rwer eo fver ygrea thel p tome . I thankD r IrP.M . Driessen,D r IrR.W .de nOuter ,D r IrIV.G .Sombroe kan d DrJ.C .Zadok s forthei radvic eo nparticula r aspects.Professo rD r IrTh.M .Worme r (Amsterdam),D r IrM .Wesse l (Wageningen),Professo rD rJ . Kranz (Giessen)an d DrH .Hindor f (Bonn)spen tman yhour so f theirtim eo nth emanuscript .Thei rcom ­ ments aboutcoffe eberr ydiseas ehav eundoubtedl y increasedth evalu eo fth e report.Part so fth eEnglis htex twer ekindl ycorrecte db yMr sSuzann eBeek , MrJ.C .Rig gan dD r IrP.M . Driessen.Assistanc ewit h theillustrativ emateria l wasprovide db yMessr sW.C.Th .Middelplaat s andG .Eimer san dthei rrespectiv e staffs.Mr sBoukjeMidden-Krosschel ltype d thevariou s stageso f thedraft ;I a m most thankfulfo rhe rcompeten twor k andefficien thelp .Mr sEll yDepryck - Steenhaardan dhe rstaf fbravel y shouldered theextr awork . 1. INTRODUCTION

Kenyacover sa nare ao f582,64 6 squarekilometer san dha sa populatio no fmor e thanthirtee nmillio npeopl e (ANON.,1976 ;MAITHA , 1976). Largelya sa conse ­ quenceo fdifference s inaltitud ean dlatitud e climatic conditions inKeny avar y markedlybetwee nareas . AsKeny a lacks sizeableminera lwealth ,th eeconom yo fth ecountr ydepend s verymuc ho nth eagricultura l sector,whic hhas ,i nadditio nt oensurin gth e majorpar to floca lfoo dconsumption ,provide d substantial exportearning san d abas efo rindustria lan dcommercia l growth (MAITHA, 1976).Th eoutpu to fap ­ proximately 90,000hectares ,plante dwit h arabicacoffee , Coffea arabica L.,i s ofparticula r economic importance,a sca nb esee nfro mth edat ai nTabl e1 (afterANON. , 1976):

Table 1:Percentage s shareso ftota lvalu eo fth eexpor to fsom emajo rcommodi ­ ties, 1970-1975 (afterANON. , 1976).

1970 1971 1972 1973 1974 1975

Coffee 21.6 18.2 20.1 22.2 18.2 16.4 Tea 12.3 11.1 13.4 10.5 9.2 10.7 Sisal 1.8 1.4 1.7 3.0 8.0 3.4 Meat and 2.8 3.5 4.0 2.3 2.1 2.3 products Other food 44.5 37.2 41. 6 39.0 34.1 33.3 and beverages

Before 1963,th eyea ro fKenya' s independence,mos to fth elarge rcoffe e plantations,coverin g approximately 31,000hectares ,wer e ownedb ynon-Africa n farmersan dforeig ncompanies .Th ecoffe eo nsmallholding swa si nAfrica nhands . Nowadaysth ecoffe e industryi st oa larg eexten tdirectl yo rindirectl y owned andmanage db yAfrican s (SENGA, 1976). The coffee industry inKeny asuffere da setbac k throughth eeconomi cde ­ pressioni nth e1930' san dthe ndurin gWorl dWa rII .Th eoccurrence ,sinc e1913 , ofth edestructiv e leafrust ,cause db y Hernileia vastatvix Berk,e tBr . (RAYNER, 1960)an dafte r192 2th ecoffe eberr ydiseas e ('CBD'),cause db y Colletotrichum

1 ooffeanum Noack,ha sadversel y affectedth eprogres so fth ecoffe e industry (MAINA, 1969).Th elatte rdiseas ewa sobserve dfo rth efirs t timei nWes t Kenya (McDONALD, 1922,1926 )an dha ssprea d fromther et oal lcoffe egrowin g areaso f Kenya,inflictin g severe crop lossesan deve nrenderin g coffee cultivationun ­ economici ncertai narea smarginall y suitedfo rcoffe e (BOCK,1970 ;FIRMA N& WALLER, 1977;RAYNER , 1952). Extensive research effortst ofin dway san dmean s ofcontrollin g CBDhav eyielde d truly effective results onlywithi nth elas t fifteenyears . Some research datao nCBD ,collecte db yth eautho ri nth eperio d 1964-196 £ have alreadybee npublishe d inscientifi can dextensio npaper s (VERMEULEN,1965 a 1966a,b ,1968* , 1970a*,b* ,c ;VERMEULE N& MEHLICH ,1966 ;VERMEULE N& PATWA , 1966*;FERNI E& VERMEULEN , 1966;HOCKIN Ge tal. , 1967*;NUBIA Ne tal. , 1968). Inth epresen tpublicatio nunpublishe d research datao flaborator yan d field experiments,includin gmycologica lan depidemiologica l studies,ar epre ­ sented togetherwit h reprintso fsom eo fth eearlie rpaper st ogiv ea genera l viewo fth eprogres s inresearc ho nth ecoffe eberr ydisease .

Reprinted in this publication 2. COFFEE CULTURE IN KENYA

2.1 Biology of Coffea spp.

2.1.1 Taxonomy

Thetaxonom yo fth egenu s Coffea (Rubiaceae)i sstil lth esubjec to fscientifi c discussions (PURSEGLOVE,1968 ;WELLMAN ,1961) .Mos tspecie so fthi sgenu sorigi ­ natei nAfric aan dthre eimportan tcommercia lspecies ,viz . Coffea arabica L. (Arabiano rarabic acoffee) , Coffea aanephora Pierree xFr .(robust acoffee )an d Coffea liberica Bull,(liberi acoffee )ar eindigenou sther e(CHEVALIER ,1947) . C. arabica isa nallotetraploi dbelieve dt ob ederive dfro mth eEas tAfrica n species Coffea eugenioides Moorean dth eWes tAfrica nspecie s C. aanephora. The distributiono fthes etw ospecie soverlap si nUganda ,whic hi sth emos tlikel y centreo forigi no f C. arabica (ROBINSON,1976) .Th ecentr eo fdomesticatio nan d variationo farabic acoffee ,however ,i si nEthiopia ,wher en oothe rwil dspecie s of Coffea occur(FIRMA N& WALLER ,1977) .

2.1.2 Varieties and cultivars

C. arabica isautogamou swit husuall yno tmor etha n9 -11 %natura loutcrossin g (CARVALHO& MONACO ,1969) .I tha stw obotanica lvarieties ,respectivel y C. arabica var. arabica and C. arabica var. bourbon (CHEVALIER,1947 ;HAARER ,1962) . Selectionwor ka tth eScot tLaboratories ,Nairobi ,ha sle dt oextensiv e plantingo fhig hyieldin gan dhig hqualit ycultivar so f C. arabica, knowna scv s S.L.2 8an dS.L .34 .

2.1.2 Morphology

2.1.S.1 General habitat

C. arabica isa relativel ysmal ltre eo rshrub ,whic hprefer stropica lhighland s (altituderang e110 0-2S0 0m )wit hcoo l(averag edail ytemperature saroun d2 4C) , butno tcol dcondition s (BROWN& COCHEME ,1969 ;HAARER ,1962) . Fig. 2: Stages of berry and bean growth (afterWORME R &NJUGUNA , 1966): a)Pinhea d stage (PH);b-c-d )expandin g stage (EXS); e-f-g)endosper m growth (ENS); h)endosper m hardening,hard-gree n berry stage (HGS); i)ripenin g fruit(RB) .

2.2 Arabica coffee in Kenya

2.2.1 Introduction, spread and importance of arabica coffee in Kenya

The cropwa s introduced toKeny a around 1893b y theearl ymissionaries .Accordin g toth eavailabl e information threedifferen t introductionswer emade ,whic hwer e known inKeny aa s 'FrenchMission ' coffee (JONES,1956 ;KIERAN ,1969 ;POWELL ,1908 / 1909; TRENCH, 1926).Mos t ofth e coffee grown inth emid-1960's ,suc ha scv s S.L. 14,2 8an d 34,originate d from these introductions. The economicpotentia l of thecoffe e industrywa s realized ata nearl y stage andwhe nwhit e settlerswer e encouraged to farm inKenya ,larg earea swer e planted with thiscrop .Befor e 1933th e cultivationo fcoffe eb yAfrican swa s prohibited foreconomi can dphytosanitar y reasons (HEYER& WAWERU , 1976).Afte r 1933per ­ missionwa sgrante d toplan t coffee ona ver y limited scale inEmbu ,Mer uan d Kisii,bu ti twa sno tunti lth een do fth e1940' stha tcoffe e growingb yAfrican s was allowedi nothe r areas (SMITH, 1976).Whe nthe ywer e actively encouragedt o participate inth ecoffe e industry (SWYNNERTON,1954 )th erol eo fth eAfrica n smallholders inth etota l coffeeproductio n increased rapidly,a sca nb esee n fromth eplante d areaan dproductio n datai nTabl e 3.Nowaday s about 270,000 African farmers,organize di ncooperativ e societies,participat e inth ecoffe e industryi nKeny a (ANON., 1973).

Table 3:Plante d area (in thousands of hectares)an d production (inthousan d metric tons) incoffe e estates and smallholdings in theperio d 1935-1974 (after ANON., 1976;KRU G &D E POERCK, 1968;SENGA , 1976).

Estates Smallholdings planted area production planted area production

1935 48.0 n.a. 0.4 n.a. 1950 24.0 6.3 0.7 0.2 1955 27.0 12.2 3.1 0.3 1960 28.3 19.1 13.3 4.7 1965 32.5 23.1 50.0 16.2 1970 29.7 27.9 62.6 30.4 1974 28.5 30.8 63.0 39.4 n.a.= no t available.

Initially,coffe ewa sgrow naroun dNairob i only,bu tgraduall y cultivation spread intoarea s aroundKiambu ,Ruiru ,Thika ,Nyeri ,Kitale ,Koru ,Soti kan d eventually eveni nth eTait aHills ,nea rVo i(ADERO , 1969;HILL ,1956 ;KRU G& DEPOERCK , 1968).Th ecoffe e areasi nWes t Kenya,viz .Kisii ,Kitale ,Koru , Kakamega,etc. ,ar egenerall y referredt oa s 'Westo fth eRift 'areas ,whil eth e regions easto fth eRif tValle y sucha sKiambu ,Thika ,Ruiru ,Embu ,Meru ,Nyeri , Limuruan dMachako sar ecalle d 'Easto fth eRift 'areas .Th e'Grea tRift 'i spar t ofth egeologica l systemo ffault s running throughKeny a roughly fromNN Et oSSW , extending fromth eU.S.S.R .t oth eSouther npar to fAfrica .Th elocation so f currentlyproducin g coffee areasar eshow ni nFig .3 .Fo reac hprovinc eth etota l areaplante dwit hcoffe ei sgive ni nTabl e4 . •**i+ International Boundary Limitso fRif tValloytapproximateonly ) ::'•'•"••'For m lints ^^ Arabica coffro areas

Fig. 3:Mai n arabica coffee areas inKenya . Table 4:Tota l estate and smallholders coffee areas in thousands ofhectare s in 1970b yprovinc e (afterANON. , 1973).

Province Nyanza Western RiftVa l ley Central Nairobi Eastern Coast

Estates • 6.1 19.8 1. 6 2.2 Small­ 10.4 4.2 0.3 27.0 19.1 1. 6 holders

Total 10.4 4.2 6.4 46.8 1. 6 21. 3 1. 6

2.2.2 Environmental and cultural aspects of coffee growing in Kenya

2.2.2.1 Environment a.Climat e

C. arabica isgrow ni nKeny ai narea swit haltitude s ranging from 120 0m u pt o 2300m an dwit h annualprecipitatio n figures from75 0m mt omor e than200 0m mpe r annum.Bot h rainfallan dtemperatur e areclosel y associatedwit h altitude:wit h increasing altitude temperatures decreasean drainfal l increases.Accordin gt o BROWN& COCHEM E (1969)arabic a coffeeunde r Kenyan conditionswil lno tthriv ei n areas whereth etemperature s regularly riseabov e 27°Co rfal lbelo w1 0C a t leastfo rman y daysi nsuccession .I nTabl e5 climatologica l datao fsom eloca ­ tionsar epresented .Th erainfal lpattern so fth emai ncoffe earea s (Fig.3 )ar e markedly affectedb yth evariation s inaltitude ,exposur ean dgeographica lposi ­ tiono r characteristics (BROWN& COCHEME , 1969;LUMB , 1966). Inmos tWes to fth e Rift areas,wit hth eexceptio no fKisi ian dSoti ki nth esouther npart ,ther ei s anabsenc eo fa distinc tdr yseaso nan donl yon erain yperiod ,th etropica lrain ­ fallpattern .I nth eEas to fth eRif ta distinc tbimoda l rainfallpattern ,th e equatorial regime,occur swit hth emai n rainyperio di nMarc ht oMa y('lon grains' ) andth eslightl y lessnoticeabl e Octobert oDecembe r rainperio d ('shortrains' ) (BROWN& COCHEME , 1969;HUXLE Ye tal. , 1969).Th erainfal lpattern sar eo fgrea t importancet ocoffee ,a sth eonse to fth erain s triggersth eflowerin gan dhenc e determinesth etim eo fth efruiting . Duet othei rbimoda l rainfall patternth ecoffe e regionsEas to fth eRif t usually experiencetw oflowerin g flushes,yieldin g respectively the'lat ecrop' , harvestedi nOctobe rt oNovembe ran dth e'earl ycrop' ,harveste di nMa yt oJune . Asa consequenc e coffee treesi ntha t region carryoverlappin g cropsa tvariou s stages.ofdevelopment .Additiona l flowering flushesar etriggere db yrai nshowers , whichar eno trar edurin gth edr yperio d fromAugus tt oSeptember . Table 5:Climatologica ldat a of some locations incoffe e areas ofKeny a (after BROWN &COCHEME , 1969;HINDORF , 1973b).

Approx. Temperature; Approx. , „,, ,r. , annual meanmonthl y _ . ._ . . altitude ., , , .. , J General rainfall regime rainfall minima and inm . inm m maxima

West of theRif t Kitale 1900 1130 10°/13°- Monomodal (Tropical) 23°/29°C regime with onerain y period Kakamega 1554 1922 14°/18°- AsKital e 26°/30°C Koru 1730 1840 10°/13°- AsKital e 23°/29C Kisii 1981 1300 n.a. Bimodal (Equatorial). Allmonth s humid (R> E0), except July Sotik 1646 1000 n.a. Similar toKisii ,bu t dry periodsmor eap ­ parent

Rift Valley

Nakuru 1840 840 7°/10°- Monomodal rain regime 24°/28°C distinct dryperio d

East of theRif t

Nyeri 1800 890 n.a. Bimodal pattern Machakos 1800 950 n.a. Bimodal pattern;distinc t (coffee areas) dry period Kiambu (Upper) 1800 830 7°/ll°- Bimodal pattern;inter - 26°/30°C mittant raindurin g dry period Embu 1680 1020 n.a. AsNyer i (coffee areas)

Meru 1850 2270 AsNyeri ;mai n rainfall inOctobe r toDecembe r n.a.= no tavailable . b.Soil s

Thesoil sprevailin gi nth ecoffe earea sbot ho fth eEas to fth eRif t(eas tan d southo fMoun tKeny a-Nairobi )an dth eWes to fth eRif t(Moun tElgon ,Kisii )ar e

10 dark reddishbrow nt odusk yre d('chocolate 'coloured) ,ver ydeep ,porous ,heav y claysbu twit ha loam y feeling.The yar ederive d from trachyteo rphonolit e lavas or tuffs,o rTertiar yo fEarl y Pleistocene age,ofte nwit ha recen t admixtureo f volcanicash .A sa grou p theyhav ebee ncalle d 'KikuyuRe dLoam' .I nth eFAO / Unesco soil legendterminolog y (1974)the ybelon gt oth eNitosols ,i nth eU.S . 'SoilTaxonomy 'classificatio n (SoilSurve yStaff ,1975 )largel yt oth eRhodi c subgroupso fth ePaleudult so rPaleudalfs ,o rt oPalehumults . Most coffeegrowin g takesplac eo nsoil sderive d frompre-Cambria ncrystal ­ lineBasemen t Complex rocks,bot hi nth eEas tRif t (Machakos)an dth eWes tRif t (Bungoma,Kakemega ,Kapenguria) .The yar ea heterogeneou s collectiono fre dan d brownsoil so fvariou sdepth ,porosit yan dtexture .Th emajorit y isclassifie da s FerricAcrisol so ra sOrthi c Ferralsols (FAO/Unesco, 1974),o ra sOrthoxi c Tropudultso rth eTropepti can dUlti cHaplortho xi nth eU.S . 'SoilTaxonomy ' system (SOMBROEK, 1979). Arabica coffee requiresa deep ,rich ,slightl y acid (pHbetwee n 5-6),fre e draining soil.Whe nth esoi l depthi sno tth elimitin g factor,th eroot so fcof ­ feear eabl et openetrat et o3-3. 5m .Th ecultivatio no fcoffe ei sconfine dt o flatarea san dt ohillsid e slopes,whic har eonl y slightlyo rmoderatel y steep. The steepestslope swit hth ehighes t hazardso ferosio nar ecarefull y protected andth efores tvegetatio ni smaintaine d (BROWN& COCHFJE , 1969).

2.2,2.2 Aspects of coffee cultivation

a.Prunin g

Twoprunin g systemsar eapplie di nKenya ,viz .th esingle-ste m system (onemai n vertical stemwit hbearin ghorizonta l shoots,secondarie san dtertiaries ,selecte d during pruningo na framewor ko fpermanen t branches)an dth emultiple-ste m system (fruits are-borno nprimar yan dsecondar y branches,produce d laterally from severalmai nvertica l stems,whic har ereplace dafte r fivet oseve n years). Inth e1950' sman y farmers adoptedth emultiple-ste m system,a si twa s allegedt ogiv ehighe ryield s (HAARER,1962 )an dgiv e lessprunin gproblems .Th e single-stem systemwa sgenerall y recommendedfo rshade dcoffee ;multiple-ste m undercondition swithou t shade (ANON., 1964).

11 b. Planting distance

Theusua lplantin g distanceunde radequat e rainfallconditions ,viz .100 0-120 0m m perannum ,i sapproximatel y 2.5x2.5m .Unde rdrie rcondition sth etree sar e planted fartherapart .I nsmallholde r areasth ecoffe e treesar eofte nplante d at2.0x2. 0m .I nplantatio ncoffe eth einterro wspacin gi susuall yno tles s than 3m t oallo wth eus eo ftractor-draw nmachiner y (WALLIS& WORMER , 1971). c.Mulc han dfertilizer s

Theapplicatio no fmulchin gmateria lha sa markedl y favourableeffec to nth e coffeeyield san dquality ,whil e soilerosio nan dwee dgrowt hdecrease .A s mulchmaterial sbanan aan dmaiz e trash,elephan tgras s (Penn-isetum purpureum) and Guinea grass (Panicum spp.)ar eapplie d (ANON., 1964). Extensive researcho nth e fertilizerrequirement so fcoffe eha syielde d impressive increasesi ncoffe epro ­ duction (ANON.,1964 )an despeciall ynitroge ndressing sar ever y rewarding (JONES et al., 1961;PEREIR A& JONES , 1956). d.Wee d control

Weed controli ncoffe eplantation s isimportant ,a scoffe e treesar equickl yaf ­ fectedb ycompetin gweeds ,especiall yi ndrie rareas .Bot hhan dweedin gan d chemical controlar eapplied ;th eforme rmainl yi nsmallholdings ,th elatte ri n plantationcoffe e (ANON., 1964). e. Irrigation

Incertai nstage so fth eberr ygrowth ,viz .th egrowt ho fth eendosper m (2.1.3.2), anadequat e supplyo fwate ri srequire dt oobtai noptima lyield s (WORMER, 1964). By irrigationthes eyield sca nb erealized .O nirrigate d fieldsa furthe r increase canb eobtaine db yreducin gth eplan tdistance s (ANON.,1964 ;WALLI S& WORMER , 1971). f.Shad e trees

Inth ehighe raltitud earea si nKeny acoffe ei sgrow nunde ra canop yo fshad e trees.Usuall y Grevillea robusta A.Cunn .i sused ,bu tals ospecie so f Acacia, Albizzia and Primus are planted.

Thesecultura l techniques,whe napplie di nKeny ab yskille dfarmers ,gav e considerableyiel d increases,especiall yi ncombinatio nwit heffectiv epes tan d diseasecontro lmethod s (2.2.4) andth e'tonic ' sprays.Th elatte r increase leaf retentionan dfoliage ,thu s augmenting coffeeyield s (RAYNER, 1957;RAYNE R& JONES, 1948).

12 2.2.3 Organization of coffee research

Before194 4coffe e researchwa scarrie dou tb yth eDepartmen to fAgriculture , whoseactivitie swer emainl yaime da tth elarge-scal eagricultura lenterprises . Althoughhampere db ylac ko ffunds ,especiall ydurin gth eeconomi c depression inth e1930's ,an d lacko fman-power ,th eresult sobtaine dwer econsiderable ,e.g . thehig hyieldin gan dqualit y cultivarsS.L .14 ,2 8an d3 4(JONES ,1956 ;RAYNER , 1952). In194 4th eKeny aGovernmen tbough tJacarand aEstat enea rRuir ua sa centrefo rcoffe e research.A ssuc hi tbecam efull yoperationa la tth een do f 1949 (ANON., 1971). UntilOctobe r196 3coffe eresearc hresorte dunde rth e Government,althoug hi treceive dfinancia l supportfro mth eCoffe eBoar do fKenya . Since theni tha sbee nru nb yth eCoffe eResearc hFoundatio n (C.R.F.),which re ­ ceivedit sfinance sfro mth eCoffe eBoar do fKeny aan dfro mth eKeny aGovernmen t (ANON., 1971).Th emai nresearc hstatio na tRuir uan dth esubstations ,locate da t Meru,Kor uan dKisii ,ar eadequatel ystaffe dan dequippe d (ANON., 1973).

2.2.4 General disease and pest situation in Kenya

Arabicacoffe esuffers ,a sha sbee nstate dbefore ,fro mtw omajo rfunga ldiseases , viz. leafrus t H. vastatrix, andcoffe eberr ydiseas e (CBD), C. aoffeanwn. Leaf rusti seffectivel ycontrolle db ypre-rai nspray so fcoppe rcontainin g fungicides (BOCK,1962a ,b ;RAYNER , 1962).Althoug hcontro lmeasure sar eeffectivel y applied, CBD stillconstitute sa grav edange r (ANON.,1975 ;FIRMA N& WALLER , 1977). A listo fdisease san dpest saffectin garabic acoffe ei nKeny ai spresente d inTabl e6 . To reducelosse scause db ypest san ddiseases ,a wid erang eo fcontro l measuresi srecommende dan dregularl yamende dt oinclud emor e effective chemicals orbette rcultura lmethod s (ANON.,1964 , 1970).Thu scoffe efarmer sar eprovide d withup-to-dat einformatio nprepare db yth estaf fo fth eCoffe eResearc hStatio n orthroug hpublication si n'Keny aCoffee' .

2.3 Terminology

Inthi san di nmos tothe rpublication so ncoffe eth ecoffe e fruitsar ecalle d 'berries',thoug hbotanicall yspeakin gth efruit sar edrupe s (=stonefruits) .As , however,th eendocar po fthi sstonefrui ti sno tver ydistinctive ,th efrui ti s called 'berry'whe nstil lgree nan d'cherry 'whe nth ecolou rchange sfro mgree n tore dupo nripening .

13 Table6 :Los to fsom emajo r diseases (A)an dpest s (B)o farabic a coffeei nKenya .

Commonnam e Causal organism Reference

A. Stem Pitting Sheffield,196 3 Warty disease Botrytis cinerea Pers.e xFr . Anon., 1970 Rootro t Armillaria mellea (Vahl.exFr. )Kumm.Anon. , 1970 Elgon die-back Pseudomonaseyringae v.Hal l Ramos& Shavdia , 1976 Fusarium bark disease Fusarium stilboides w. Siddiqi, 1965 Leaf blightan dste m Phoma tarda (Stew.)comb . nov. Boerema& Bollen , 1975; dieback (basionym Asaoohyta tarda) Firman, 1965; Steward, 1957 (p. 430). Leaf rust Hemileia vastatrix Berk,e tB r Rayner,196 0 Coffee berry disease Colletotriahum aoffeanumNoac k Firman& Waller , 1977 Berry blotch Ceraospora aoffeicola Berk & CookeVa nde rVosse n Cook, 1975

Coffee leafmine r Ghes.an d Leucoptera meyriaki } Crowe,196 4 L. caffeina Washb. Antestia Antestiopsis lineatiaollis Stal. Wheatley,196 2 Thrips Diarthrothrips aoffeae Will. Anon.,197 0 Mealy bugs Planoaoaousspp . Anon.,197 0 Stem borers Anthores leuoonotus Pasc. Tapley, 1960 Dirphya nigriaornis 01. Crowe, 1959;Anon . > 1970 Bixadus sierriaota White } Anon., 1970 Euaosma nereidopaMeyr . < Lace-wing coffeebu g Anon.,197 0 Habroahila ghesquierei Sch. Capsid bugs Anon.,197 0 Lamproaapsidea coffeae Ch. Loopers Crowe& Leeuwangh , 1965 Epigynopteryx strictigramma Hmps. Ascotts selenaria reoiproaaria Wlk.Wheatley ,196 4 Berry borer Hypothenemus hampei-Ferr . } Anon., 1970 Berrymot h Phophantis smaragdinaButl . Leaf skeletonizer Leuaoplema dohertyi Warren Crowe, 1960 Brown scale Saissetia aoffeae Wlk. Anon., 1970 Star scale Asteroleoanium aoffeae Newst. Crowe, 1962

14 With regard toth e anatomical aspects discussed inthi spaper , 'bark maturation'wil l oftenb ementioned . Inyoun g twigs ofcoffe ea phelloge nde ­ velops inth ecorte x ofgree ninternode s inconjunctio nwit h increased radial growtho fth ebar k tissues.Thi sphelloge nproduce s the firstperider m orsecon ­ darycork. A serieso fperiderm s issuccessivel y formedbeneat h the firston ea s theaxi scontinue s toincreas e incircumference .Thi sproces s of secondary cork formationan d thesubsequen t colourchange s ofth e surfacetissue s ofth e coffee twigs is generally referred toa s 'barkmaturation' . Inthi spublicatio n the termsestate s and smallholdings areuse d asfollows : a.estates :typica l commercial coffeeenterprises ,usuall y runan downe db ynon - Africans and inmos t cases financially capable ofcarryin g outexpensiv eagrono ­ mical andplan tprotectio nmeasures ;b .smallholdings :smal l coffeeplanting s ina systemwhic h ist o asmalle ro rlarge rexten tpar to fsubsistenc eagriculture , ownedb yAfricans ,operatin g through cooperative societies and oftenfinanciall y restricted toadher e strictly toagronomica l andplan tprotectio nrecommendations , required for themaintenanc e ofth ecoffe etrees . The following abbreviations areuse d regularly:a.s.l .= abovese alevel ; CBD =coffe eberr ydisease ;CR F= Coffe e Research Foundation;CR S= Coffe e Research Station;cv(s )= cultivar(s);S.C .= sporulatin g capacity;S.L . =Scot tLabora ­ tories.

15 3. LITERATURE REVIEW OF COFFEE BERRY DISEASE

3.1 Discoveryan dsprea do fcoffe e berry disease inKeny a

Coffeeberr ydiseas e (CBD)wa sdetecte dfo rth efirs ttim ei n192 2nea rSo yan d Turboi nWester nKeny aclos et oth eborde rwit hUgand a (McDONALD, 1922).Th e disease spreadsouth -an deastward san db yth elat e 1930'sal lhig haltitud e coffeearea sWes to fth eRif twer eaffecte d (RAYNER, 1952). In193 9CB Dwa sfoun do na fe wcoffe eplantation snea rNyer io nth eslope s ofMoun tKeny a (Fig.3) .Th ediseas eappeare di nth ecoffe egrowin g district aroundKiamb ui n1951 ,still ,however ,restricte dt oaltitude s above approximately 1700m (RAYNER, 1952).I nth elat e 1950'sCB Dha dsprea dthroug hal lhig haltitud e coffeearea sEas to fth eRif t (NUTMAN& ROBERTS ,1960a , b). Coffeearea sbelo w 1700m altitudewer efoun dt ob eaffecte di n196 2an da t theen do fth e1960' sal lcoffe egrowin gdistrict si nKeny a (Fig.3 )suffere d seriouscro p losses (GRIFFITHS, 1970).

3.2 Economic losses inKeny a

McDONALD (1926)estimate dtha tCBD ,i nth efirs tyear so fit sappearance ,cause d incertai nsmal larea so fWes tKeny acro p losseso fapproximatel y 75°s.I nth e earlyan dmiddl e 1930'sCB Dinduce dcro p lossesi nth ecoffe e areasWes to fth e Rift,amountin g oftent o25 %o fth eexpecte dyield s (RAYNER, 1952); insom e marginalarea scoffe egrowin gwa sabandone di nfavou ro fothe rcas hcrop s (BOCK, 1970). Theoutbreak so fth ediseas ei nth emai ncoffe earea so fth eCentra lProvinc e resultedafte r195 1i nsom e instancesi nyiel dreductions ,whic hwer eestimate db y BOCK (1963)a tsom e7 5-80 1o fth enorma lyield .NUTMA N (1966)reckon stha tCB D reducedth epotentia lcro pi n196 4wit h approximately 191.MOG K& HINDOR F (1975) recordedi nthei robservatio nsite si nth eKiamb uare ayiel dlosse so fu pt o801 . Theinformatio navailabl e (ANON.,1973 )suggest s thatth eyield so festat ean d smallholding coffeedroppe dfro mrespectivel y80 0an d100 0k gpe rhectar et ono t muchmor etha n40 0k gpe rhectar ebecaus eo fCBD .Sinc ethe nth eestat eyield s have recoveredt opre-196 6levels ,bu tth esmallholde ryield shav eremaine dabou t 400t o50 0k gpe rhectar ebecaus eth ecost so fCB Dcontro lar eofte nprohibitiv e forth esmallholders . 16 3.3 Symptomso fcoffe e berry disease

Flowerbudsan dflower sar eparticularl y susceptiblet oCBD .Th efruit s exhibita n increased susceptibility duringcertai n stageso fdevelopmen t(Figs,2 ,6 )particular ­ lyth eexpandin g stagean dpar to fth eendosper m stageo fth esoft-gree nberr y andth eripenin gcherry .The yca nals ob eaffecte d duringth epinhea dan dhard - berry stages (MOGK,1970 ;MOG K& HINDORF , 1975;MULINGE , 1970a). The first symptomso fattac kar eusuall y dark-brownblotche so nflowerbud s and flowerso rstreak so nth ewhit epetals .Th elesion s increase rapidlyi nsiz e anddestructio no fth eflowe ri scomplet ewithi n4 8hours .O nberries ,especiall y expanding berries,th efirs t symptomsar esmal ldark-brow nspots ,ofte no nth e lateral surfaceo fth eberry .Thes e lesions enlargei narea ,becom e slightly sunkenan deventuall ycove rth ewhol eberry ,whil eth eberr y contentsar eals oin ­ vaded.Affecte dberrie s either fall (berriesi nth eearl y stageso fdevelopment ) or remaino nth etre e (hard-berry stage)a sblackish ,mummifie d berries.CB Dca n also causesymptom so nleave san dstems ,a sshow nb yHINDOR F (1973c).Thes ewil l notb ediscusse di nthi spublication . Inth esunke nlesion sminute ,slightl yprojectin g punctuations develop,whic h areth efruitin g bodies (acervuli)o fth efungus .Unde rdam pcondition s theydevel ­ opmasse so fpinkis hconidia .I fth edevelopmen to fthes e lesions ishalte db yth e onseto fdr yweathe r conditions,the ychang e ofteni ncolou r frombrown-blac kt o ashen-grey,dotte dwit hblac kacervuli .Thi s typeo flesion sha sbee n termed 'scab'b yMcDONAL D (1932)an di tindicat eth einactiv epresenc eo fCB Do ngree n berries. 'Scab'lesion sma ychang e intoactiv e (brown-black) lesions againwhe n damp conditionsar ere-established . When ripeningcherries ,whic har ever y susceptiblet oCBD ,ar eaffecte d yield losses tendt ob erathe r limited,bu tth eprocessin go fthes e affectedberrie sma y giveconsiderabl epulpin gproblems .Thi s stageo fth ediseas e isofte nreferre dt o asth e'brow nblight ' stage,know nalread yi nEas tAfric abefor eth eappearanc eo f CBD (McDONALD, 1921;SMALL , 1926)an da ttha ttim eprobabl yno tassociate dwit h C. aoffeanwn (McDONALD,1926) .

3.4 Occurrenceo fcoffe e berry disease outside Kenya

Exceptfo ron erecen t reporto nth eincidenc eo fCB Di nBrazil* ,th ediseas eha s

* Dr P. Figueiredo (Inst.Biol , di Sao Paulo,Brazil) ,pers .comm .t o DrD .Mulde r (Wageningen)

17 onlybee nreporte dfro marabic agrowin gcountrie si nAfrica :Tanzani a(TAPLEY , 1964);Ethiopi a(FERNIE ,1966 ;GASSERT ,1976 ,1978 ;MULINGE ,1975) ;Ugand a(BUT T &BUTTERS ,1966) ;Rwand a(FOUCAR T& BRION ,1963) ;Zair e(HENRARD ,1957) ;Ivor y Coast(BOISSON ,1960 ;MEIFFREN ,1957) ;Cameroo n(MULLER ,1964 ,1970 ,1973 ,1978 ; MULLER& GESTIN ,1967 )an dAngol a(D APONTE ,1966) .I nth eCentra lAfrica n Republic,SACCA San dCHARPENTIE R(1969a ,b )foun d C. ooffeanum alsoo nberries , twigsan dleave so f C. oanephova (robustacoffee )an d Coffea dewevrei deWil d& Dur. (excelsacoffee) .

3.5 Colletotriohvm spp.o ncoffe e

Colletotriohvm spp.hav ebee nassociate dwit hvariou ssymptom so ncoffee ,viz . theanthracnos esymptom so nleaves ,branche san dfruit san dth edie-bac ksymptom s ofbranches ,characterize db yth edesiccatio no fbranc htip s(BOISSON ,1960 ; BUTLER,1918 ;HENRARD ,1957 ;HOCKING ,1966 ;MuTHAPPA ,1970 ;SACCA S& CHARPENTIER , 1969a,b ;SMALL ,1926) .Mos to fthes e Colletotriohum isolatesar eclassifie da s C. ooffeanum, afterth eisolatio ndescribe db yNOAC Ki n190 1fro mcoffe ei n Brazil,an dar econidia lstage so f Glomerella aingulata (Stonem.)Sp .& Schr . SMALL(1926 )alread ypointe dou ttha t Colletotriohvm conidiaar es ocommonl y foundo ncoffee ,tha ta larg eproportio no fhealth ygree ntwig san dleave syielde d acervulio f Colletotriohum, whenincubate dunde rmois tconditions .RAYNE R(1948 ) postulatedtha tth etranslucen tyello wlea fspots ,know na s 'weakspots' ,wer e causedb ya laten tinfectio no f Colletotriohvm. Hefoun dtha tpracticall you to f everypiec e (upt o1 m m) o fhealth yleaf ,gree nberr yste man dpedice la rang e ofdifferen t Colletotriohum spp.coul db eobtaine dafte rsurfac esterilizatio nan d subsequentplatin go nagar .CRITCHET T(1969 )presente dals oevidenc efo ra possi ­ blelaten tphas eo fCB Di ngree ncoffe eberries .SHA W(1967 )reporte da diffuse , yellowlea fspo to narabic acoffe ei nPapu aNe wGuinea ,yieldin ga Gloeosporium sp.an da G. aingulata andi twa ssuggeste dtha tth efungu swa spresen ta sa latentinfection . Theubiquitou soccurrenc eo f Colletotriohum spp.particularl yo ncoffee ,bu t alsoo nman yothe rtropica lplants ,a sreporte db yworker sthroughou tth eworl d (e.g.BOISSON ,1960 ;BUTLER ,1918 ;FERNANDEZ ,1961 ;HENRARD ,1957 ;HINDORF ,1975 ; HINDORF& MuTHAPPA ,1974 ;HOCKING ,1966 ;MEIFFREN ,1957 ;MUTHAPPA ,1970 ;SIMMONDS , 1963; VARGAS& GONZALES ,1972 ;WELLMAN ,1954 )ha srecentl ybee nfurthe rdocumente d byth einformatio ngathere db yMULDE R (personnalcommunication ,1978 )i ncooper ­ ationwit hresearc hworker si nSout h andCentra lAfrica ncoffe egrowin gcountries . Numerous Colletotriohum isolatesfro mdifferen tpart so fcoffe etree si nthes e countrieswer esen tt oth eNetherland sfo rscreenin go nth epresenc eo f C. coffeanum. Hence,i ti sno tsurprisin gtha t Colletotriohurn spp.ar eofte nisolate d fromcoffe ewit ha rang eo fsymptom so rphysiologica ldisorder s(necroti cspots , moribundtissue ,branche swit hdie-bac ksymptoms) .A nexampl efro mKeny ai sth e Elgondieback ,attribute dt o C. ooffeanum (NUTMAN& ROBERTS ,1960a) ,bu tcause d by Pseudomonas syringae v.Hal l (RAMOS& SHAVDIA ,1976) .O nth eothe rhan di ti s certaintha tth elaten tinfectio no fleave sassociate dwit hprematur elea ffal l (RAYNER,1948 )ca nb econtrolle dwit h 'tonic'spray so fcoppe rfungicides .Thi s suggestsa relationshi pwit hth e Colletotriohurn sp.isolate dfro mth elea fstal k (HOCKING,1967a ;HOLLIES ,1967 ;RAYNER ,1957 ;RAYNE R& JONES ,1948) .

3.6Taxonom yo f Colletotriohurn spp.

Thetaxonomi cclassificatio no f Colletotriohurn speciesfoun do ncoffe ei sstil l confusedan dunsettled .Th e C. ooffeanum describedb yNOAC K(1901 )wa sno ta pathogeno fgree nberrie san dth enam eha sbee nuse dfo rpracticall yal l Colletotriohurn isolatesobtaine dfro mcoffee ,regardles swhethe rsaprophyti c orparasiti c (FIRMAN& WALLER ,1977 ;SMALL ,1926) . InEas tAfric aMcDONAL D(1926 )an dRAYNE R (1941,1948 )alread ynotice d morphologicaldifference si nthei r Colletotriohurn isolatesfro mcoffe ean dth e latterdescribe dthes edifference sclearly ,proposin gth enam e C. ooffeanum var. virulans forth eCB Dfor misolate dfro mgree nberrie s (RAYNER,1941 ,1952) . MEIFFREN (1957)i nth eIvor yCoas trecognize dthre e Colletotriohurn strainsbase d onconidia lsiz ean dsetae .NUTMA N& ROBERT S(1960b )considere dth eCB Dpathoge n morphologicallyindistinguishabl efro mth enon-pathogeni cforms ,a sdescribe db y SMALL (1926). In196 9GIBB Sdifferentiate dth eisolate so f Colletotriohurn obtainedfro m coffee.H eclassifie dfou rmai n 'strains'o f Colletotriohurn bycolon ycharacter ­ istics,usin gsingl espor eisolates :CB D( = C. ooffeanum var. virulans, RAYNER, 1941), oop (= C. ooffeanum, 'pink')., com( = C. coffeanum, mycelialform) ,an d ooa (= C. ooffeanum, acervuliform) .Fo rth efirs ttim ea clea rdifferentiatio n hadbee nmad ebetwee nth e Colletotriohurn speciesinhabitin gcoffe ebar kan d hithertoconsidere dt ob e C. ooffeanum, thecaus eo fcoffe eberr ydiseas e (NUTMAN& ROBERTS ,1961) . Later,HINDOR F (1970,1972 ,1973a ,b )mad edetaile dmorphologica lstudie s ofth eisolate so f Colletotriohurn, obtainedfro mal lpart so fth ecoffe etree .B y discriminantanalysi so f in vitro measurementso frange so fmorphologica l

19 characteristicsan drelatin ghi sfinding st oVO NARX' swor k (1957,1970) ,h ecam e toth efollowin gclassification : Colletotriohum aoffeanvm Noack {sensu stricto) =CB D C. acutatum Simm.(GIBBS ,1969 : ccp) C. gloeosporioid.es Penz.(GIBBS ,1969 : cam) C. gloeosyorioides (GIBBS,1969 : oca) C. gloeosporioides =greenis hmycelia lfor m(VERMEULEM ,1970b ) Glomerella eingulata (Stonem.)Sp .& Schr .= whit esteril emycelium ,peritheci a after2- 3week s Thethre e C. gloeosporioides isolatesar econidia lstage so f G. eingulata. Thegreenis hmycelia lform , C. gloeosporioides, wasals ofoun di nTanzani ao naf ­ fectedberrie s (JOHANNS,cite db yHINDORF ,1972) . Althoughi ti srealize dtha tth eHINDOR Fclassificatio nha sdistinc ttaxo - nomicaldrawbacks ,i ti sstil lconsidere dth emos tsuitabl esyste mavailabl ea s itprovide sa nacceptabl emean sfo rth edelimitatio no fth evariou scomponent so f the Colletotriohum complex.

3.7Othe rhos tplant so fcoffe eberr ydiseas e

SMALL(1926 )wa sth efirs tt ocarr you tcros sinoculatio nexperiment swit h Colletotriohum isolates.Thes eisolate swer eobtaine dfro mcoffe ean dothe rcul ­ tivatedplant si nUganda .Unde rcertai ncondition sth ecoffe eisolate scoul din ­ fectothe rhost san d vice versa. Theresult swere ,however ,erratic . HOCKING (1971)carrie dou tinfectio ntest so nth efruit so fplant so fnine ­ teendifferen tfamilies ,usin ga CB Disolat ean da Colletotriohum isolatefro ma coffeetwig .Accordin gt othi sautho rh ecoul dinfec tth efruit so fthre eplan t specieswit hth eCB Disolate .Wit hth etwi gisolat eh ewa sabl et oinduc eactiv e lesionso nfourtee no fth eninetee nspecies .CRITCHET T (1969)reporte do nth e behaviouro f C. aoffeanvm onothe rtropica lfruits .H einduce donc ediseas esymp ­ tomsb yinoculatio no fa suspensio no f C. coffeanum conidiao ntomat ofruits . SMALL,HOCKIN Gan dCRITCHET Ttrie dt oinfec t in vitro otherplant so rfruit s ofthes eplants .Thei rresult swer eo nth ewhol eno tver yconclusive .Th eiso ­ lationo f Colletotriohum spp.b yHINDOR F(1974 )fro m2 8hos tplant si nKeny aof ­ feredth eopportunit yt otes tth epathogenicit yo fthes eisolate so ncoffe e berriesan dt ocompar eth ecultura lcharacteristic swit hth eCB Dfungu sfro m coffee.Non eo fth eisolate scoul dinfec tgree nberrie so rwer emorphologicall y similart o C. coffeanum. HINDORF (1974)foun dfurthermore ,tha t C. coffeanum in­ oculatedo na numbe ro fplant scoul dno tinduc ean ysymptoms ,bu ttha tsapro ­ phytic C. gloeosporioides and G. eingulata wereno thos tspecific .Thes eresult s

20 elucidateth efinding so fSMAL L (1926)an dHOCKIN G (1971). Apart fromth eallege d incidenceo f C. coffeanwn onothe r Coffea spp. (SACCAS& CHARPENTIER , 1969a,b )an dth eisolatio no f C. coffeanum fromth e barko fth eshad etre e Sesbania sesban (L.)Meril li nEthiopi a (GASSERT, 1978), no furtherevidenc ei sye tavailabl eo nth epresenc eo fCB Do nothe rhos tplants .

3.8CB Dresearc h inKeny a

3.8.1 Research between 1922 and 1950

In192 2CB Dwa sfoun di nWes tKeny a (McDONALD, 1922).Accordin gt osom e farmers thediseas eha dalread ybee npresen tfo rsom eyear s (RAYNER,1952 )an da si t spread throughth ehig haltitud earea so fWes tKenya ,researc heffort swer e ini­ tially focussedo nth enatur eo fth epathogen . McDONALD (1926)foun dtha ta for mo f Colletotrichum coffeanum Noackwa s al­ ways associatedwit hth ediseas e symptomso ngree nberries .Morphologicall y this formcoul db eeasil ydistinguishe d fromthos eknow nt ob eassociate dwit hblight , die-backo fbranche san dlea fsymptom s (BUTLER,1918 ;McDONALD ,1921 ;SMALL , 1926). Inserie so fexperiments ,McDONAL D (1925,1930 , 1931, 1932,1936 )studie d such aspectsa sinoculation , in vitro culture,viabilit yo fconidi aan dtoxicit yo f copperan doi lspray st oconidia ,bu tth eresult so fhi sexperiment swer ein ­ conclusive. Througha numbe ro fquestionnaire s sentt ocoffe e farmersi nth eaffecte d areas,i tcoul db eascertaine d thatCB Dwa sespeciall y severei ncomparativel y wet regionswit hhig hrelativ ehumidities .Furthermor eth elo wtemperature so fth e highaltitud eareas ,i ncombinatio nwit hhig hrelativ ehumidity ,wer e foundt oco ­ incidewit hhig h levelso fCBD .On eo fth emos tcommo ntheorie sa ttha ttim ewa s the 'weakened tree'concept ,whic hblame dth econditio no ftree si nneglecte d plantationsfo rthei rhig hCB Dsusceptibility .Prope rwee d control,remova lan d destructiono fdea dbranche san dtrees ,correc tprunin gan dapplicatio no flim eo r calcium carbideBordeau xspray swer erecommende d (McDONALD,1926 ;RAYNER , 1952). The theorytha tonl yneglecte d treeswer e susceptiblet oCB Dcoul dno tb esub ­ stantiated,a sextensiv e field trialswit hmanur ean dfertilizer s carriedou t throughoutWes tKeny afaile dt oprov ean yrelationshi pbetwee nlo wlevel so f majoro rmino rnutrient san dhig hlevel so fCB D(RAYNER , 1952). Itwa sfoun d thatfungicid espray si ngenera lexerte da favourabl eeffec to n thetrees ,viz .b yextendin g leafretention ,resultin gi ndense r foliage.Thi sef ­ fectwa stwent yyear slate rattribute d especiallyt ocoppe rfungicide s (RAYNER,1957 ;

21 RAYNER& JONES , 1948). Extensive fungicide screening trials,carrie d out inman y parts ofWes tKenya ,di dno t resulti nan y clear recommendation fora specifi c fungicide:mos to fth efungicide s applied seemed ablet osom edegre eo fcontrol ­ lingCBD .RAYNE R (1952)pointe d out that frequent sprayswoul dprovid ea nalmos t complete fungicide covero nth edevelopin g fruitswhic hwoul d result inbette rCB D control.Thi sobservatio no fRAYNE Rwa s ignoredi nth e late 1950'san d early 1960's,bu twa s completelyvindicate da tth een do fth e 1960'swhe n itwa s clearly shown that fungicidal sprays applieda sprotectant s during therain swer eeffec ­ tivet ocontro l CBD (GRIFFITHSe tal. , 1971;MULLER , 1964,1968 ,1970 ,1973 ,1978; MULLER& GESTIN , 1967). Inth eperio d 1922-195 0th e onlyprogres si nth e fieldo fCB D controlwa s theselectio no fth ediseas e resistant cultivarBlu eMountai nb yth eAgricultura l Department. At theen do fth e 1950'sRAYNE R (1952)summarize d theresearc h achievements and listed the following researchpriorities : -a ninvestigatio no fth eparasitis mo fal l formso rstrain so f Colletotriohum oncoffe e and theeffect so fth etissues ,i nwhic h these forms develop,o nth e morphologyo f Colletotr-ichwn; - themetho db ywhic h the infectionpasse s fromon e cropt oanother ; -th e effecto fclimat eo nth edisease ; -th eeffec to fth ephysiolog yo fth etre eo nit s susceptibilityt oth edisease ; -th e causes'ofresistanc e and thesubsequen t search for resistanthig hyielding , high qualitycultivars . With this listo fprioritie s RAYNERprovide da ttha t timealread ya clea r outlineo fth eproblem st ob esolved .

3.8.2 Research between 1950 and 1966

After 1951,CB D spread through themajo r coffeearea si nth eCentra l Province, but no effective controlmeasure swer ereadil y availablet oreduc e the losses caused by thesene woutbreaks . Itwa s realized thatne wresearc h shouldb einitiate d without delay. As earlya s195 6BOC Kprovide d alreadyvaluabl e informationo nfiel dan dla ­ boratory inoculum techniques,thu s enablinga standardizatio no finoculatio nex ­ periments.BOC K& RAYNE R (1956)achieve d effective field controlo fCB Dwit h monthly sprays ofpheny lmercury-acetat e during therain yperiods ,combine dwit h fortnightly sprayso fcoppe r fungicides.Becaus eo fth ephytotoxi c effectso f mercuryo ncoffe e and thepossibilit y ofmercur y residuesi nbean s (BOCKe t al.,

22 1958)thi spractic e couldno tb erecommende d forapplicatio n incommercia l coffee plantings.Th e search formor e effective fungicideswa s continued (BOCK, 1959)an d eventually itwa s established that copper fungicides gaveth ebes t control ofCB D whenapplie d before theonse to fth erain s (BOCK, 1963;CLARK E &WILLIAMSON ,1963) . Inth emeantime ,NUTMA N& ROBERT S (1960a,b , 1961)studie d factors affecting conidial germination and infectionb y thepathoge nan dth erelationshi p of these factors toth ediseas edistribution .The y found thatth e fungus capableo finfec ­ ting greencoffe eberrie s inhabited thebar ko fcoffe ebranches ,wher e itha dre ­ placed the saprophytic Colletotriohum. Thematurin gbar k thusprovided , according toNUTMA N& ROBERT S (1961),a continuou s source of conidial inoculum responsible forberr y infectionan d thepositio no fmaximu m inoculumpotentia l coincidedver y closelywit h theregio no fmaximu m crop concentrations.Hence ,i twa s concluded that thediseas e level isapparentl y a functiono fth edegre e towhic h thema ­ turingbar k isinfested .Thi s theorybecam eknow na s the 'inoculumpotential ' theory.The y suggested furthermore that theCB Dpathoge nma yhav e arisenb ymu ­ tation froma mildl yparasiti c straino f C. ooffeanvm atsom e timeprio r to 1922. Both the germination of C. ooffeanvm conidiaan d the infectiono fth eberrie s cantak eplac ewithi n thetemperatur e range of 17 -2 8C ,a swa s establishedb y NUTMAN andROBERT S (1960b).A combination ofsuitabl e temperatures and favourable highrelativ ehumidit y fora perio d of fivehour swoul d allow infectiono fth e cropan dthes e sharplydefine d conditionswoul d explain therestrictio n ofth e disease tohig h altitude areas (NUTMAN &ROBERTS , 1960a,b) .The y suggested that valuable diseaseprediction s couldb ederive d frommeteorologica l datainterpre ­ tation (NUTMAN& ROBERTS ,1960b , 1969a). The 'inoculumpotential 'theor ywa s generallyaccepted ,especiall ywhe nBOC K (1963),CLARK E andWILLIAMSO N (1963)an dNUTMA N andROBERT S (1962a)coul ddemon ­ strate theeffectivenes s of copper fungicides,applie dbefor e theonse to fth e rains.Hig hvolum e sprays ofcoppe r fungicides (NUTMAN &ROBERTS ,1962a )applie d atth etim eo fmaximu m inoculumpotentia l ofth epathoge n justbefor e ordurin g the onseto f therains ,woul d reduce the inoculumpotentia l to low levels and hence reduce the infectiono fberries. Thealread ymentione d 'tonic'effect s ofcoppe ran d the fact that copper fungicideswer e alsoeffectiv e against leafrus t contributed greatly toth ein ­ creased application ofcoppe r formulations inth e late 1950'san dth eearl y 1960's. Thusfar theselectio no fCB Dresistan t coffee cultivars didno t appear tob e verypromising . FIRMAN (1964)screene d over 100specie s andcultivar s and found a number ofthes e still tob e free from infectionafte ra perio d of twoyears . However,hi smateria l didno tposses s theeconomicall y interesting characteristics

23 ofhig hqualit y andyield .Moreover ,a schemica l control seemedt ob eeffective , further concentrated researcheffort si nthi s fieldwer eno tpursue da ttha ttime . Inth ebeginnin go fth e 1960'sth eCB Dproble m seemedt ohav ebee nsolve d and researcheffort swer e directedt oothe rdiseas eproblems ,e.g . leafrust .Afte r 1962,whe nCB Dals o spread into loweraltitud earea s (below 1700m a.s.l.),i t be­ came increasingly evident that copper fungicides applieda tth erecommende d time, volume andpressur ewer e stillno tefficacious .I n1964 ,however ,th e lossesdu e toCB Dbecam e calamitous.WALLI San dFIRMA N (1965,1967 )foun d that lowervolumes , at the samequantit yo fcoppe r compoundpe runi t area,gav ea simila r levelo fdis ­ ease controla sth erecommende d highvolume s (NUBIAN& ROBERTS , 1962a).WALLI San d FIRMAN (1967)wer e also the firstt ofin da discrepanc y between the inoculumpoten ­ tial datacollecte d routinelyi nthei rtrial s and theactua ldiseas e incidence. FIRMAN (1965)briefl y reviewed theachievement s after 1950an d listed research approaches deemedurgent : -th emechanis mb ywhic h the inoculumpotentia lo fth ebearin gwoo di sreduce db y fungicidalsprays ; -a stud yo fth ecompetitiv e colonizationo fth ematurin g barkb yth evariou s formso f Cotletotriohum; - therelationshi p between growthan dmaturit yo fth ebranche s andth etimin go f fungicidal sprays,s otha t timing canb erelate dt obiologica l criteria instead ofcalende rdates .

3.8.3 Research after 1966

Following reports ofmountin g lossesa sa resul to fth econtinuou s spreado fCB D and the apparent inadequacyo fth erecommende d chemical controlmeasures ,researc h effortswer e renewedi n1964 . Considering the importanceo fCB D forth ewhol e arabica industryi nEas tAfrica ,NUTMA Nan dROBERT Swer eposte da tth eEas t AfricanAgricultura l andForestr y ResearchOrganizatio n (E.A.A.F.R.O.)t oin ­ vestigate- i ncooperatio nwit hth ecoffe e research stationso fth e threeEas t African countries (Kenya,Ugand a andTanzania) - th ebiolog y andecolog yo fth e pathogeni ndetail .Th eCoffe e Research Stationa tRuir uconcentrate d before 1966 onth e search formor e effective fungicides.Thi s researchwa s carriedou ti nco ­ operationwit ha numbe ro fchemica lmanufacturers ,wh o- a tthi s stage- provide d fungicides only.Later ,thes emanufacturer s also actively carried out screening trialso fthei row n (e.g.COLLINS ,1968 ;FORDYC E& SHAW ,1968 ;BAYER ,cite db y VERMEULEN, 1968).Afte r 1966th enumbe ro fresearc hworker s engagedi nCB Dre ­ search throughout Kenya,increase d considerably through thesuppor to fth eBritish ,

24 West Germanan dNetherland s governments. In 1964/1965i twa s stillbelieve db yman ypeopl e thatth esprea d ofCB D into loweraltitud e areaswa smainl y causedb y achang e ofth e climate triggered by the floods in1961/196 2 (NUTMAN& ROBERTS ,1962a ,1969a ,d) .Consequentl y the established control schedules sucha spre-rai n sprays,wer e recommended tob eap ­ plied evenmor evigorousl y (WALLIS,1964 )a sn obasi c changesha dbee nmad e inth e disease controlconcept .Th echange dweathe r conceptwa s laterrefute db y BROWN andCOCHEM E (1969)an dGRIFFITH S andWALLE R (1971).Th e latterworker s analyzed rainfall data from 1951 onwards,whe nCB D firstappeare d inth emai nEas t Rift districts and they couldno t findan y evidence thatth e climatic conditions during the 1960'sdiffere d from thoseo fth e 1950's,neithe r inth ehig hno r inth e low altitudecoffe eareas . Intensive screening trialso f fungicideswer e carried outa tth eCoffe e ResearchStation ,base d on theconcep to fpre-rai n sprays (BOCK, 1963;CLARK E& WILLIAMSON, 1963).Th epreliminar y results ofthes e trialswer epresente d atth e First SpecialistMeetin g onCoffe eResearc h inEas tAfrica ,hel d inNairob i in 1966 (VERMEULEN, 1966b).Th e finaloutcom eo fthes eexperiment s indicated thedis ­ ease controlling effectso fOrth oDifolata n (captafol)a scompare d toth e standard copper formulations (VERMEULEN, 1968). Inthes escreenin g trials itwa s also found that the inoculumpotentia l data,routinel y assessed according toth e technique describedb yNUTMA N andROBERT S (1961),di dno t coincidewit h the actual incidence ofth ediseas e onberries ,especiall y towards theen do fth etrials .Thi s infor­ mationwas ,however ,no t considered sufficient to invalidate the inoculumpoten ­ tial theory.Moreover ,a compound ,viz .Tuzet ,wa s founda t thattim ewhic hre ­ duced the inoculumpotentia l sodrastically ,tha t itwa sbelieve d thatth ecur e toCB Dha dbee n found (NUTMAN &ROBERTS , 1969b). At theaforementione d First SpecialistMeetin g onCoffe eResearc ha tNairobi , MILLER reported onhi swor k onCB D inth eCameroon .H eha d achieved efficient controlo fCB Db yrepeate d applications ofcoppe r fungicides during thesingl e rainy season inth eCameroon , thusprovidin g aprotectio n for the expanding berries (MULLER, 1964).MULLE R suggested a similarapproac h forKenya ,bu thi s suggestionswer e considered inapplicable toKenya ,du e todifference s inclimat e and croppingpattern ,althoug hVERMEULE N andPATW A (1966)reporte d favourablere ­ sultsb yextende dspra yapplication s atMer uan dKitale . Ina subsequen tvisit , MILLER (1968)furthe rdetaile dhi s suggestionswit h specialreferenc e to thecon ­ ditions inKenya . Inth emeantime ,mor e timing trialsha d alreadybee n carried out andver y sooni twa s discovered thatMULLER' s suggestions indeedprove d tob eef ­ fective (GRIFFITHS &GIBBS ,1969 ;GRIFFITH Se t al., 1971).Th e recommendations

25 wereamende daccordingly ,usin gOrth oDifolata nan d later forsom etim eals o Benlatea sth emos teffectiv e fungicides (VINEe tal. , 1973a, b). Inth esecon dhal fo fth e 1960'sconsiderabl e advanceswer emad e invariou s fieldso fCB Dresearch .Whe nGIBB S (1969)manage d todifferentiat e thecomponent s ofth e Colletotrichum complex inhabiting thebar ko fbearin gbranches ,thi sclassi ­ ficationforme d thebasi s fordetaile dwor kb yHINDOR F (1970,1973a ,b ,c) , MULINGE (1971a,b )an dVERMEULE N (1970a,b) .Th econidia lproductio nharveste d frombearin gbranche swa s found toconsis tmainl yo fsaprophyti c conidiao f Colletotrichum. Thesefinding s led toa completel y different evaluationo fth e inoculumpotentia l theorya sdraw nu pb yNUTMA Nan dROBERT S (1960a,b ,1961 , 1969a)an dsuccessfull yuse db yBOC K (1963). FURTADO (1969,1970 )an dVERMEULE N (1970b,c )investigate d theeffect so f copper fungicides onth ecompositio no fth e Colletotrichum complex.The y found thati nspraye dcoffe e fieldsa shif t towards C. ooffeanvm incidenceoccurre dwith ­ inth ecomple xan dthi sresulte d ina nincrease d levelo fth epathogen .Thi sshif t inth e Colletotrichum populationha dbee nalread y thesubjec to fhypothetica lcon ­ siderationsb yROBINSO Nan dWALLE R (1966)an dMULDE Ran dHOCKIN G (1967),whil ea t a laterstag e theeffect so ffungicide sno tonl yo nCBD ,bu tals oo n leafrust , foliationan dyield ,wer e furtherdiscusse d (FIRMAN, 1970;GRIFFITHS ,1971 ; HINDORF, 1973d;MULING E &GRIFFITHS ,1974a ;NUTMA N& ROBERTS , 1969c).Th erelatio n betweensusceptibilit y toCB Dan dcro pdevelopmen t stages (MOGK,1970 ;MOG K& HINDORF,1975 ;MULINGE , 1970a), thedispersio no fconidi ao f C. ooffeanvm intree s (MULINGE,1970b ,1971a ;WALLER , 1972a),th eeffect so faltitud eo n C. coffeanum distribution (HINDORF,1972 ,1974 ;MULINGE ,1971b) , thetemperatur e andrai ncon ­ ditionsnecessar y forCB Dinfectio nan dfungicida l effectso ndiseas edevelopmen t (MOGK, 1973;MULING E &GRIFFITHS ,1974b ;STEINER , 1973a,b ,c , 1974)wer eamon g theothe rresearc h findings,whic hcontribute dgreatl y toth eknowledg e ofCBD . Although theconcerte d effortsaime da tth ecausa lorganis mo fCB Dan d its controlme twit h considerable success,i twa s realized thati nth elon gru nthi s approachwoul d offera tbes tonl ya partia ldiseas econtro lan d certainlyb eo fa temporarynature ,considerin g thetarge to fth edisease ,viz .th eexpandin gber ­ riesan d thephysica l locationo fth epathoge ni nth ematurin gbark .Therefor e an extensive coffeebreedin gprogramm ewa s initiated, financedb yth eGovernment s of Kenyaan d theNetherlands .Thi sbreedin gprogramm e stillcontinue s andth eresult s regardinghig hqualit yan d yielding cultivarswit hCB D resistant characteristics aremos tencouraging ,involvin gals o international cooperationbetwee ncoffe epro ­ ducing countries (VANDE RVOSSEN , 1973,1974 ;VA NDE RVOSSE Ne tal. ,1976) . Atpresen t allaforementione dexpatriat e research staffhav ebee nwithdrawn ,

26 with theexceptio no fth eplan tbreede rstil lresidin g atth eCoffe eResearc h Stationa tRuiru .Curren t researcho nCB D islargel y geared toth e screening of fungicideswit h special reference toth e increased fungalresistanc e to systemic fungicides (BAKER, 1972, 1973;COOK , 1975;COO K &PEREIRA , 1977;OKIOGA ,1976 ; OKIOGA& MULINGE , 1974). FIRMAN andWALLE R (1977)hav e recently reviewed indetai l all researchen ­ deavours since 1922.

27 4. MATERIALS AND METHODS

4.1Coffe e cultivars

Inlaborator yexperiment s detached,green ,expandin gberrie s (MOGK,1970 ;MOG K& HINDORF,1975 ;MULINGE ,1970a )o fth eCB Dsusceptibl e cultivarsS.L .2 8an d3 4 (JONES,1956 )wer euse dfo rth eassessmen to fpathogenicit yan dfo rconidia lger ­ minationtrials . Fordetail so ncultivar suse di nfungicid e screeningan dspra y timingtrial s reference shouldb emad et oth ereprinte dpaper sb yVERMEULEN ,1968 ;p .80 ) and VERMEULENan dPATW A (1966;p .66) .Th esam eapplie sfo rinformatio no nth e cultivarsuse di nth eanatomica lan dmycologica lstudie s (VERMEULEN, 1970a,b ; p. 36,p . 44). The fieldobservation so naspect so fCB Do nAfrica nsmallholdin g siteswer e carriedou to nth ecultivar sS.L .34 , FrenchMissio nan dKent' sselectio nK7 .

4.2Fung i

Forth elaborator y screeningo ffungicides ,conidi ao fth eCB Dpathogen , C. coffeanum werea tfirs tharveste d fromCB Dlesion so ncherrie so fcv .S.L .3 4 asdetaile di nth ereprinte dpape r (VERMEULEN,1968 ;p .80) .Whe ni twa sfoun d thato nrip echerrie sothe rsaprophyti c Colletotviahvm spp.wer e alsopresen t (GIBBS, 1969),onl ygree nberrie swit hmarke dCB Dlesion swer eused ,bu totherwis e theroutin eremaine dth esame . Inth epathogenicit yan dgerminatio ntest sconidia l suspensionso fth evari ­ ous Colletotrichvm spp.an do f G. oingulata obtained frommono-spor e cultures (TUITE,1969 )ar eliste di nth ereprinte dpape r (VERMEULEN,1970a ;p . 36).

4.3Fungicide s

A listo fchemical san dfungicide steste di nlaborator yan dfiel dexperiment si s presentedi nTabl e7 . Cadmium containing chemicalswer e includeda sstandar dchemical si nth elabo ­ ratorytests ,becaus e cadmiumi spositione dnea rmercur y inth eperiodi c system andmigh tposses sth esam estron gCB Dcontrollin g characteristicsa smercur ywith ­ out,however ,th ephytotoxi can dresidua l side-effects.Moreover ,cadmiu m

28 Table 7:Chemica l and tradenames ,cod enumber s andmanufacturer s of the compounds screened for CBD controlling characteristics.

Chemical nameo r codenumbe r Trade name Manufacturer Remarks

cadmium chloride crystals - Pure cadmium sulphide - chemicals cadmium sulphate mercury bichloride - - used as standard stannous citrate

cuprous oxide Perenox I.C.I. Shell-75 - Shell Code number cuprammonium carbonate Fungex Murphy Chem. cupric oxychloride Coprantol I.C.I. copper oxychloride20 % Colloidox Agricola cupric sulphate Cupravit-Blu e Bayer fentinhydroxide Du-Ter Philips-Duphar fentinacetate Brestan Hoechst anilazine Dyrene ChemagroCo . glyodin Crag Fruit Union Carbide Fungicide 341 \ cycloheximide Actidione UpjohnCo . blasticidin-S Blasticidin Shell ethylenebis (tetra- Fungicide 328 DuPon t Experimental hydrodimethyl thiadiazine- fungicide thione) captan Orthocide Chevron Co. EL-211 - Eli Lilly Code number lead arsenate Lead arsenate - zineb +nicke l ethylene Sabithane-M Rohm &Haa sCo . bisdithiocarbamate mancozeb DithaneM 45 Rohm &Haa sCo . dodine acetate Cuprex Am.Cyanami d zineb + ferbam +mane b Tricarbamix Vondelingenplaat folpet (Ortho) ChevronCo . Phaltan captafol (Ortho) Chevron Co. Difolatan chlorothalonil Daconil Diamond Co. (DAC 2787) Olin-1763 - OlinCo . Code Olin-1562 - OlinCo . numbers dichlorophen Halophen Sindar (Turbicide) venturicidin 10% Venturicidin Murphy Chem. dichloran Allisan Boots Pure DrugCo . phenyl mercury-acetate Verdasan I.C.I. maneb Manzate-D DuPon t ziram Zerlate Rohm &Haa sCo .

29 fungicides arealread ycommerciall yavailable .Stannou s citratewa s includedt o provide astandar d forth eti ncompound s included inth e listo ffungicides .

4.4 Isolation offung i

Themethods ,use dfo risolatio no ffung i fromplan tmaterial ,ar edescribe d inth e reprintedpaper s (VERMEULEN,1970a ,b ;p . 36,p .44) .

4.5 Inoculationmethod s

Theassessmen t ofconidia lpathogenicit y ofisolate so rmixe d suspensions of Colletotriohum spp. (4.2)wa s carriedou to nexpanding ,gree nberrie s (4.1), following the inoculationtechniqu e ofBOC K (1956).T o further standardize this technique,droplet so f0.02 5m lo fconidia l suspensions ofknow n concentrations and compositionwer epipette dwit h amicropipe t onto theberr ysurfaces . Theboxes ,eac hcontainin g 100berrie s andwe tblottin gpaper ,wer ekep t closed for 15-20 days.Startin g fromth e 8thda y after inoculational l infected berries,viz .wit hCB D lesions originating fromth e inoculationdroplet ,wer ere ­ moved andcounte dever y secondday .Th epercentag e ofaffecte dberrie s after 18-20 dayswa s adopteda sa measur e ofth epathogenicit y ofth einoculate d conidial sus­ pension. Inoculationtest swer e alsocarrie dou twit hmixe d conidial suspensions of C. aoffeanum andeac ho f thethre esaprophyti c Colletotriohum spp. (HINDORF,1970 , 1973a)a tconcentratio n levelsvaryin g from 10t o 10 conidiape rml .Fo reac h concentration level fourmixtur e combinations ofpathoge n: saprophyt ewer e tested: 4:1, 3:2, 2: 3 an d 1: 4 . Mixedconidia l suspensions,similarl yobtaine d fromth e incubated andsub ­ sequentlywashed•branche s (NUTMAN& ROBERTS ,1961 )wer e inoculated onberrie sa s describedunde r 4.7.

4.6 Assessment of conidial germination and pathogenicity

Theexperiment s included germination tests fora rang eo fconidia l concentrations of C. aoffeanum, thesaprophyti c Colletotriohum spp.an dmixture so fconidi ao f Colletotriohum spp. (3.6,4.5 ) inhangin g dropso fsteril ewate r inVa nTieghe m cells,o nslide swit hwate r agar (Fig.4 )an dberr y surfaces.Furthermor e theef ­ fectswer e investigatedo fconidia l exudateso nth egerminatio nan dpathogenicit y of C. aoffeanum onaga ran do nberrie s respectively,whil e alsoth epathogenicit y

30 agarcoatin g

cigarettepaper ; conidialsuspensio n

Fig. 4:Slide s coated withwate r agaruse d for germination tests (A)an d to establish effects of exudates (B)o f Colletotriohum spp. of C. ooffeanum alone and inmixture s ofconidia l suspensionswit h the saprophytic Colletotriohum spp.o nberrie swa s assessed.Conidi ao fth e Colletotriohum spp. were obtained asdescribe dunde r 4.2. Germinationrate swer e observed after8-1 0 hours incubationa ta constan t temperature of 22°C (NUTMAN& ROBERTS , 1960b); germinationwa s considered positive whenth egerminatio n tubewa s atleas t 2\ x the lengtho fth econidium ,o rwhe na n appressoriumha dbee n formed. The effect ofconidia l exudateso ngerminatio no f C. ooffeanum onaga rwa s assessedb y seeding suspensions (0.1m l droplets)o fconidi a (10 perml )o fa saprophytic Colletotriohum sp.o ncigarett epape rplace d onon ehal fo fa divide d blocko faga r (Fig.4) .Th epape rwit h the suspensiono nto po f itwa s removed the nextmorning ,takin g careno t tospil l the suspension.Subsequentl y onbot hhalve s ofth eaga rblock s asuspensio no f C. ooffeanum conidiawa s thenpipette d (0.1m l dropletso f 10 conidia/ml).Afte r 8-10 hours cottonblue-lactophenol was added toth eaga rblock s andth e germination assessedunde r themicroscope . The effecto f theexudate s ofth esaprophyti c Colletotriohum spp.o nth e germination ofan d the infectionb y C. ooffeanum ofberrie swa s investigated by seeding droplets (0.02Sml )o fconidia l suspensions (10 conidia/ml)o n cigarette paper disks (1|c mdiameter )place d onto po fgree nberries .Th enex tmornin g

31 thesedisk s andth e suspensions onthe mwer e removed and theberrie s immediately inoculatedwit hdroplet s (0.025ml )o fa CB D fungus conidia suspension (5x1 0 conidia/ml).Th e infection ratewa s assessed inth e samewa y asdescribe d by BOCK (1956). Wheno nberrie s only the germinationwa s assessed theboxe swit h the berries were opened themornin g after inoculation; the inoculation dropletswer e rapidly driedunde r afa nan da thincoatin g ofcolourles s nail-varnish applied to the dried-up droplet.Afte r someminute s the coating could bestrippe d offeasil y and mounted incottonblue-lactophenol . Clear castso fth eberr y surfacewit h theger ­ minating conidiawer e thus obtained and countswer eb emad eunde r themicroscope . 3 7 For thehighe r concentrations (10 -1 0 conidia/ml) thegerminatio n of 100conidi a * inever ydrople twa s scored, inth eorde r inwhic h they came intoview . Germination and subsequent infection onth e surface ofberrie s ofgree nCB D susceptible cultivarswa s assessed by inoculating expanding berrieswit h a range of concentrations of theCB D fungus andals o conidiao f C. aoffeanwn inth e four mixture combinations and concentration levels (4.5)wit h theconidi ao fth e saprophytic Colletotriohum spp.

4.7 Production of conidia on thebar k and their pathogenicity

For the fungicide screening trials,th e sporulating capacity of the Colletotriohum population incoffe ebar k (GIBBS,1969 ;GRIFFITH S &GIBBS ,1969 )wa s assessedac ­ cording to theNUTMA Nan dROBERTS 'techniqu e (1961, 1969a). Slight modifications of this techniquewer euse d for theobservatio n inth eAfrica n smallholding areas. Random samples of 25branche spe r sitewer e collected atthree-weekl y and later at five-weekly intervals.A t thedista l endo feac hbranc h thegree ninter - nodes (Fig. 5)wer e discarded, thecu tbein gmad e one internodeabov e thatwhic h showed theobviou s first signs ofbar kmaturation .Th e surface areao fth einter - nodeswit h little orn o signs ofbar kmaturatio nwa s still smoothan d uncracked by cork formation.Th ebasa lportion s ofeac hbranc h (Fig. 5)wer e also discarded, leaving aneight-internod e lengtho ftwi g (NUTMAN &ROBERTS , 1961).Al l these lengthswer e divided intosingl e internodeswit h cuts atth enode s and subsequent­ lyplace d into screw-topped jars labelled 7-14, the first containing all the 25 distal internodes ofth ebranche s fromon esit e andth e lastthos ewhic hwer e fullymatur e (Fig. 5).Th e sub-sampleswer e thenwashe d intw o changes ofwate r by shaking them ina mechanica l shaker for fiveminutes .Afte rdraining ,th e bottleswer e incubated for 24hour s at 22-24C .Th enewl yproduce d conidia onth e brancheswer e thenrecovere d from the internodes by gently rinsing themateria l in

32 ,firs t signo f barkmaturation numbero f 9 tO 11 12 13 K 15 16 17 18 internode

greeninternode s fungus presenti nthi s area true bark

Fig. 5:Longitudina l schematic presentation of a coffee branch and thevariou s stages of barkmaturation .Th e internodes arenumbere d from theape x to thebasa l part of thebranch .Bar kmaturatio n usually starts at inter­ node 8,bu t thisma y bedifferen t fromon e branch toanother .

twoseparat echange so fwate r (10m leach )whic hwer esubsequentl y combinedt oon e sampleo f2 0ml .Hal fo fthi s2 0m lsuspensio nwa sthe ncentrifuged ;th econidi a inth epellet swer ekille dan dstaine db yadditio no ftw odroplet so faci d fuchsin inlacto-phenol ,an dth evolume sbrough tu pt o1 ml .Th econidi ai nthes esample s werecounte di nduplicat eo na haemocytometer .Th esurfac eare ao fth einternode s wasdetermine d frommeasurement so fthei rlengt han dwate rdisplacement .Th e numbero fconidi aproduce d fromeac htwenty-fiv e internodebatc hwa sexpresse d 2 asth econidi aproductio npe rc m perhour . However,i norde rt oobtai na necessar yreductio ni nth evolum eo fth ework , fromApri l 1968onward sth einternode s 7+8 ,9+10 , 11+1 2an d13+1 4ha dt ob e takentogether .Otherwis eth eprocedur eremaine dth esame . Theothe rhal fo fth e2 0m lconidia lsuspensio nwa suse dfo rpathogenicit y testso ngree nberrie s (4.5). Carewa stake nt oadjus tth esuspension so fal l samplesb ydilutin go rcentrifugin gt oa concentratio no f2x1 0 conidia/mli n ordert ocompar eth einfectivit yo fal lsuspension swit heac hothe ran dwit hth e infectivityo fa 2x1 0 conidia/ml suspensiono ffres h C. ooffeanum conidia.Thes e inoculationtest swer ecarrie dou tfro mMarc h 1967t oJun e 1969.

33 4.8Assessmen t ofCB Dlevel s infiel d experiments

Asi tprove d difficultt oge treliabl e yielddat aan dthu st oobtain -informatio n onth eeffect so fCB Do nyiel d fromth eowner so fsmallholdin g sites,visua las ­ sessmentso nth eactua lCB Dincidenc ewer emad eregularl yfo reac hsit ea tEmb u andMeru .Te nbranche spe rsit ewer e selecteda trando man dmarked .Th efollowin g scoring systemwa sapplied ,usin g onlyth enumber so fvisuall y affected flowers and fruits,excludin gth eintensit yo fth einfection :

noCB Dpresen t 1- 20% CBD 21- 401CB D 41- 60%CB D 61- 801CB D 81- 100 %CB D absenceo fflowers , pinheads and/or berries

Considering thevaryin g levelso fsusceptibilit y toCB D(MOGK ,1970 ;MOG K& HINDORF, 1975;MULINGE , 1970a)i nth enatura l sequence (Table2 ,Fig .6 )o f flowers (FL)- pinhead s (PH)-soft-gree nberrie s (EXS+ ENS ) -hard-gree n berries (HGS)-rip e berries (KB),a scor eo f*-* -2-1- 0 wouldmean ,tha t approximately 21-40 1o fth esoft-gree nberrie san d1- 20 1o fth ehard-gree nberrie sha dbee n affectedb yCB Dan dtha tn oflower san dpinhead sha dbee npresent .A scor eo f 1_*_*_1^2woul dmean ,tha t 1-20 1o fth eflowers ,1 -20 %o fth ehard-gree n berries and 21-40 %o fth erip eberrie sha dbee naffecte db yCBD ;pinhead san dsoft-gree n berriesha dbee nabsent .

Susceptible Non-susceptible Susceptible stage Non- Susceptible stage stage _ stage stage Pinhead Hard-green Flowering stage Soft-green berry stage berry stage. Ripening stage (PH) Expansion stage : Endosperm stage (HGS) berries 7WT (EXS) 4ESSL 0 2 4 G 8 10 12 14 16 18 20 22 24 26 28 30 32 34 WEEKS AFTER FLOWERING

Fig. 6:Natura l sequence of flowering and cropping stages inconjunctio n with susceptibility toCB D (afterMOGK , 1970;WORME R &NJUGUNA , 1966).

34 4.9 Screening of fungicides

The laboratory and field screeningmethods,use d for screening of fungicides, are listed inth e reprintedpape r on fungicide screening (VERMEULEN, 1968;p .80) .

35 5. ANATOMICAL AND MYCOLOGICAL ASPECTS

5.1 Coffee berry disease in Kenya. I. Colletotrichum spp. colonizing the bark of Coffea arabica

H.VERMEULEN 1

East African Agriculture and Forestry Research Organization, P.O.B. 30148, Nairobi, Kenya

Accepted 2 April 1970

Abstract

Several species of Colletotrichum occur in maturing bark of Coffea arabica branches in Kenya. The Colletotrichumpopulatio n inhabits thebar k tissueexterna l to thedevelopin g phellogensi nth ecortex . The Colletotrichum species are unable to invade green bark tissue, where the phellogen has not yet been differentiated, whilecolonizatio n ceases on thephelloder m of thetru e bark. Only oneo f the Colletotrichum speciesdiscusse di nthi spaper , C.coffeanum, can infect green coffee berries.Shortl y after the initiation of thefirst phelloge n inth ecorte x theparasit e isi na small areai n the bark. It cannot befoun d in bark tissues where more phellogens have been formed and where the colour of the bark haschange d from yellow-green to brown or black. From the bark of C.liberica trees,grow n inKenya , and C. arabicacv . 'Bourbon', grown in green­ housesi nth eNetherland s oneo f thesaprophyti ccomponent s of the Colletotrichumpopulatio n could be isolated. It wasimpossibl et oinduc edie-bac k symptomso rmer einfectio n byinoculatio n ofgree n internodes even after wounding of live branches of C. arabica with any of the Colletotrichum components colo­ nizing the bark. It is suggested that die-back systems of coffee in Kenya are primarily caused by unfavourable growing conditions.

Introduction

Colletotrichumcoffeanum Noack occurs on Coffeaarabica L. sometimes asa pathogen but mainly as a saprophyte in all coffee-growing areas of the world (Noack, 1901; Butler, 1918; Small, 1926; Gutierrez, 1954 and Meiffren, 1957). Recently Saccas and Charpentier (1969) reported the occurrence of C.coffeanum on C.canephora Pierre ex Froehner (Robusta coffee) and C.dewevreid . W. & Dur. (Excelsa coffee) in the Central African Republic. In 1922th e first definite report of a fungal attack on green coffee berries appeared onth e files of the Department of Agriculture in Kenya (Rayner, 1952). Macdonald (1926) described the causal agent as a pathogenic form of C.coffeanum. The losses due to berry infection and berry drop were serious and the disease was subsequently named coffee berry disease (CBD). With CBD gradually spreading through all high-altitude coffee-growing areas (above 1800 m.) of Kenya, research efforts were concentrated on the control of the disease (Macdonald, 1937; Rayner, 1952). These efforts, however, met with little success. Around 1951 CBD had established itself in all high-altitude areas in Kenya. A separate CBD research-unit was founded in 1955 to investigate all aspects of the disease. 1 Present address: Internationaal Agrarisch Centrum (I.A.C.), Wageningen, The Netherlands. 36 As a result of the work of the CBD research-unit, Nutman and Roberts (1960 and 1961) reported that C.coffeanum habitually colonizes the maturing bark external to the developing phellogens in the cortex. They suggested (1961) that the pathogenic form, which probably arose as a mutant before 1922, had displaced the saprophytic form of C.coffeanum i n the cortical tissues of bearing wood in heavily affected CBD areas.Thes eworker sshowe dtha t thesporulatio n ofth epathoge n on thebearin gwoo d supplied inoculum for both flowers and fruits, while under certain weather conditions the pathogen in the bark would be able to induce 'Elgon die-back' symptoms (Nut- man and Roberts, 1960).Reductio n ofthi s'inoculu m potential' (Nutman and Roberts 1961) by applying fungicidal sprays before the onset of the rains would give a better and more economic control, than protecting developing berries. Bock (1963) reported favourable results with pre-rain copper sprays. Shortly after the CBD research-unit had been dissolved and after the very wet 1961-1962 period losses due to the diseaseincrease d alarmingly when the low- altitude coffee areas (below 1800 m), previously free of the disease, became infested with CBD. New research efforts were initiated in 1964. At that time it became apparent that the recommended sprayprogramm e ofpre-rai n copper applications was partially or totally ineffective (Wallis and Firman, 1965). A reassessment of earlier work was considered necessary, also because of information obtained from other African countries where CBD occurs (Muller, 1964; Mendes da Ponte, 1966;But t and Butters, 1966). Recently Gibbs (1969) and Hindorf (1970) reported in detail on the various compo­ nents of the Colletotrichumpopulation , isolated from C.arabicai n Kenya. In view of the confused state of the taxonomy of Colletotrichumth e former author used the generic name alone for the four groups of isolates by him on coffee branches. Hin­ dorf made the following classification for his isolates, based on theircultura l charac­ teristics: 1. C.coffeanum Noack (Rayner, 1952: C.coffeanum var. virulans; Gibbs, 1969;CBD) , 2. C.acutatum Simmonds (Gibbs, 1969: ccp), 3. C.gloeosporioidesPenz y(Gibbs , 1969: ccm), 4. C.gloeosporioides(Vermeulen , 1970), 5. C.gloeosporioides(Gibbs , 1969:ccd) and 6. Glomerellacingulata (Stonem. ) Spauld. & v. Schr. In this classification isolate 1 is the causal agent of the coffee berry disease while isolates 2,3 ,5 and 6 are saprophytes. The mildly parasitic nature of isolate 4wil l be discussed at a later stage (Vermeulen, 1970). This paper will cover the anatomical aspects of bark maturation in relation to the colonization of cortical tissues by the various isolates of Colletotrichum (Rayner, 1948; Mulder and Hocking, 1967) and the sporulating capacity of bearing wood of coffee branches (Nutman and Roberts, 1961). The classification given by Hindorf (1970)wil lb euse d in this paper.

Materialsan dmethod s

To investigate the invasion of the fungus in the cortical tissues in relation to bark maturation (Nutman and Roberts, 1961) branches of three coffee varieties ('Harar', 'S.L. 34' 'Mocha'), susceptible to C.coffeanum, were cut off from trees at the National

37 first sign of 1 bark maturation

1 2 3 4 5 6 7 8 9 10 tl 12 13 V. 15 16 17 ^^r.v..»«,,v,*.to«'.hii^.M,Msi»^

green internodes fungus present in this true bark

Fig. 1. Longitudinal diagram of a coffee branch. The internodes are numbered from the apex to the basal part of the branch. The position of internodes showing the first signs of bark maturation can differ from one branch to another. Colletotrichum spp. are only present in internodes 7-14, not in green internodes or in true (fully mature) bark.

Fig. 1. Lengtediagram van een koffietak. De internodien zijn genummerd vanafde top naar het oudere gedeelte van de tak. De plaats van internodien met het eerste teken van bastrijpheid kan van tak tot tak verschillend zijn. Colletotrichum soorten zijn alleen te vinden in internodien 7-14 en niet in de groene internodien of in de internodien waar de bastrijpheid volkomen is.

Agriculture Laboratories, Nairobi, Kenya, altitude 1700 m. Three or four branches per variety were divided into internodes and placed serially into screw-top jars, each with a number indicating the position of the specific internodes on the branch and the degree of bark maturation (Fig.l), The internodes in the bottles werewashe d for three minutes in 0.1% HgCl2 and rinsed twice with sterile water. Pieces of about 1mm 2 were cut aseptically from the bark of successive internodes and plated out on water or malt agar. From each internode a series of at least seven or eight pieces of bark was taken along its length, progressing from the distal part towards the basal part of the internode. Any fungus growth originating from the bark pieces on the agar plates was transferred to malt agar after three to four days. The cultures were identified after two to three weeks The conidia of C.coffeanum-tike cultures were used for inoculation on green berries in order to test each isolate for pathogenicity (Bock, 1956). Material treated as described above was also taken from C.libericaBull , ex Hiern. and C.arabica cv. 'Bourbon' trees. The former trees were grown at the National Agriculture Laboratories, Nairobi, Kenya and the latter in a greenhouse of the De­ partment of Tropical Crop Husbandry of the Agricultural University, Wageningen, The Netherlands. Inoculation experiments were carried out with the Colletotrichum isolates 1, 2, 3 and 5 obtained from maturing bark to assess their capability to invade green inter­ nodes of live branches of Arabica trees (cv. 'S.L. 34') and induce symptoms similar to die-back symptoms (Thorold, 1945; Nutman and Roberts, 1960). Agar discs from pure cultures of each of the strains were placed on internodes, which had first been surface sterilized with 0.1% HgCl2 and washed with water. The inoculum was kept moist with wet cotton-wool and fixed with a piece of polythene sheet around the internodes. After applying the inoculum around the internodes, half of the treated internodes waswounde d by needle-pricks. Of all internodes freehand, freeze microtome and paraffin cross sections were made to study the anatomical changes in"th e cortex of the maturing bark in the course,o f the invasion by the components of the Colletotrichumpopulation . For the freehand and freeze microtome sections' cotton blue-lactophenol or Sudan IV stains were used. Theparaffi n sections were stained with thionin-light green - orange G, according to Margolena's method (Gray, 1954, p.499) .

38 Results

Nine extensive sets of experiments were carried out from March 1969 to June 1969. In all experiments attention was paid to the colour of the pieces cut from the bark material. It was found that the bark maturation progress often differed from one branch to another on the same tree. Sometimes, also, internodes were already yellow- brown or brown on top, while they were still green on the lower side of the same internode. Therefore, it was decided to use the colour of the bark as a criterion, rather than the actual position of the internode on the branch (Fig.l). In Fig.2 the pieces of bark taken from the internodes are grouped according to colour of the bark. From each group 200piece s of bark were taken, giving a total of 1200piece s of bark plated out. It can be seentha t out of the six Colletotrichum isolates defined by Hindorf (1970,) isolate 3 was found most frequently in the groups green II, yellow, yellow/brown I and brown II, followed by isolate 2 and 5. C.coffeanum, the causal agent of CBD, could be found exclusively in groups green II, yellow and yellow/brown, though it occurred even there at low frequency, In the cortex of these tissues the first phellogen hadjus t been initiated (Fig.3 )nea r the stonecell-layer, and the exterior cortical tissue had been cut off very recently from the cambial tissue by the phellogen. As soon as more phellogens had been formed (Fig.4 ) and the colour of the bark had changed from yellow-green to brown no C.coffeanum isolates could be obtained from the bark tissue. No growth of any Colletotrichumha s been detected in green bark tissues which did not show signs of phellogen formation (Fig.5) . On the true bark, where two or threephellogen s had been formed, no Colletotrichum isolates could be obtained either.

*. Fig. 2. Total of all cultures of four Colletotrichum isolates obtained in nine tests from 1200 pieces of bark cut from the area where Colletotrichum / '"••••. "•3 is present and plated out on agar. These pieces 50J of bark material are not grouped according to the •J actual internode position, but according to the colour of the bark; from each group 200 pieces i of bark were taken: green I: completely green bark tissue, no phello­ ,4 \ gen in cortex; • f i2 / --05 green II (bordering yellow bark); yellow and /' / yellow/brown: one phellogen in cortex; brown I / (bordering yellow/brown bark) and brown II: i brown bark tissue, two phellogens in cortex. —*^*r .^C.coff. C. coff. = C. coffeanum = CBD; 2 = C. acuta- *- c * i*i 6 c colour Of „ ,-, . . . t r • *-, , i- jj= s s | |- g= turn; 3 = C. gloeosponoides; 5 = C. gloeospo- S c, I .£ * * birk riQides Fig. 2. Totaal van alle cultures van vier Colletotrichum isolaten verkregen in negen proeven uit 1200 stukjes bast van het gebied waar Colletotrichum aanwezig is en uitgelegd op agarvoedingsbodem. Deze stukjes bast zijn niet ingedeeld volgens de positie van de internode op de tak, maar ingedeeld volgens de kleur van de bast; van iederegroep waren 200 stukjes bast genomen: groen I: volkomen groen bastweefsel, geen fellogeen in de cortex; groen II (liggend naast degelebast),geelengeellbruin: e'enfellogeen in de cortex; bruin I (liggend naast geel/bruine bast) enbruin II: bruin bastweefsel, twee fellogeenlagen in de cortex. C. coff. = C. coffeanum = CBD; 2 = C. acutatum; 3 = C. gloeosporioides; 5 = C. gloeospo- rioides.

39 Fig. 3. Cross section through green-yellow bark of a coffee branch. The first phellogen (P) has been formed just outside the stonecell layer (S). Magnification about x 130. Fig. 3. Dwarsdoorsnede van groen-gele bast van een koffietak. Het eerstefellogeen (P) is gevormd juist buiten de steencellaag (S). Vergroting onge- veer 130 x.

•pA^rrS^r^ t? I

Fig. 4. Cross section through brown bark of a coffee branch. Two phellogens (P) and two stone-cell layers (S) are now present in the bark. Magnification about x 130. Fig. 4. Dwarsdoorsnede van bruine bast van een koffietak. Tweefellogeenlagen (P) en twee steen- cellagen (S) zijn nu aanwezig in de bast. Vergro­ ting ongeveer 130 x.

Fig. 5. Cross section through green bark of coffee branch. Outside the cambial tissue (C) £, only one stonecell layer (S) has been formed. Magnification about x 130.

Fig. 5. Dwarsdoorsnede van groene bast van een koffietak. Buiten de cambiale laag (C) is slechts een steencellaag gevormd. Vergroting ongeveer 130 x.

40 The results of the inoculation experiments with Colletotrichum isolates 1, 2, 3 and 5o n livegree n internodes alwaysyielde d negative results and no symptoms of die-back could be induced in any of the trials even after wounding the green tissue with needle pricks. This agrees with the conclusion from the results mentioned above, that the Colletotrichum isolates are not able to invade tissue where no phellogen has been formed. The isolates made from C.liberica and C.arabica cv. 'Bourbon' yielded in all cases isolate 3. The 'Bourbon' material had been maintained in greenhouses for at least twelve years and it seems likely that the fungus was already present in the coffee cut­ tingsimporte d into the Netherlands.

Discussion

From the data presented in this paper it becomes evident that C.coffeanum could only be isolated infrequently from green II, yellow and yellow-brown bark tissues (Fig.2) , where the first phellogen had been formed. The other, saprophytic isolates are present more frequently in the same tissues (respectively approximately ten and twenty times higher than C.coffeanum), and continue to be present in the brown bark (Fig.2 , brown I and II respectively). This means that on a bearing branch the role of C.coffea­ numi nth e Colletotrichumpopulatio n isusuall y relatively unimportant and thepresen ­ ce of the parasite is virtually restricted to a small area just apical of the zone with brown colouration. The observations of Nicholls (1969) indicate, however, that there can be appreciable increase in the amount of yellow bark in the early part of the Long Rains (March-April) on certain types of branches. Considering the data present­ ed in this paper the increase of yellow bark surface would mean a higher incidence of C.coffeanum and this might explain the CBD-outbreaks during or shortly after the rains. Data presented by Gibbs (1969) and Hindorf (1970), similarly show the relati­ vely low proportion of C.coffeanum i n the Colletotrichumpopulation . The fact that C.coffeanum cannot occupy brown cortical tissue has most likely nutritional reasons. No explanation can as yet be offered for the fact that the saprophytic isolates of Colletotrichumar e also capable of occupying green II and yellow bark. The hypothesis put forward by Mulder and Hocking (1967), that the saprophytes and the parasite would occupy cell-layers at different depths in the cortical tissues might apply to the green-yellowtissues . Furthermore it becomes clear that the conidia produced on the eight internodes specified by Nutman and Roberts (1961) will be mainly conidia of the saprophytic isolates. Vermeulen(1968 )propose d the term 'potential conidial production', cover­ ing all Colletotrichum conidia produced on the branch. More recently the term 'sporu- lating capacity' (s.c.) has been introduced (Gibbs, 1969; Griffiths and Gibbs, 1969; Nutman and Roberts, 1969). With our present knowledge of the ratio saprophytic- parasitic conidia produced on the bark iti seas yt o understand whyWalli s and Firman (1965) and Vermeulen (1968) had difficulties to relate their s.c. findings with actual diseaseincidence . It should be kept in mind that the data presented in this paper apply to an altitude of approximately 1700m . Hindorf (personal communication) has observed variations in the composition of the Colletotrichum population with the altitude. The role of C.coffeanum i nth ewhol epopulatio n remains,however , relatively small at allaltitudes .

41 The failure to induce die-back symptoms in green internodes of healthy coffee bran­ cheswit h the Colletotrichumisolate s 1,2,3 and 5,eve n after wounding thebar ktissues , suggests that die-back symptoms of branches are indeed most likely to be caused primarily by unfavourable growing conditions (Thorold, 1945). Under adverse conditions the ubiquitous Colletotrichumma y invade the green bark tissues and even attack the cambial layer, thus causing die-back symptoms. No experimental evidence could be obtained that C.coffeanum is able to invade green internodes of live branches (Nutman and Roberts, 1960).

Acknowledgments

I wish to thank the Kenya Ministry of Agriculture for the help and facilities received in the period 1967-1969. This paper is published with the permission of the Director of the East African Agriculture & Forestry Research Organization, Nairobi, Kenya and the Ministry of Foreign Affairs of the Government of the Netherlands.

Samenvatting

Koffiebesziektein Kenia. I. Colletotrichum spp. diede bast vanCoffea arabica kolonise- ren.

Een aantal soorten van Colletotrichum komt voor in de rijpende bast van Coffea arabica takken (Fig. 1 en 2) in Kenya. De schimmelpopulatie koloniseert het bast- weefsel, dat zich in de cortex buiten de ontwikkelende fellogeenlagen (Fig.3 en 4) bevindt. Deze schimmels zijn echter niet in staat groen bastweefsel te koloniseren, waar nog geen fellogeen (Fig.5 )i s gevormd. In het deel van de tak waar echte schors voorkomt, met in de cortex twee tot drie fellogeenlagen, kon geen Colletotrichum worden gei'soleerd. De Colletotrichum soort, die groene koffiebessen kan infecteren, C.coffeanum, is slechts sporadisch te vinden in een klein gebied van de bast kort na de vorming van het eerste fellogeen (Fig.3 ) in de cortex. In bastweefsel waar meer fellogeenlagen zijn aangelegd en de kleur van de bast van groen-geel tot bruin of zwart isveranderd , komt C.coffeanum niet meervoor , maarwe lander eColletotrichum soorten (Fig.4) . Met geen van de Colletotrichum soorten - beschreven in dit artikel - konden instervings-symptomen worden geinduceerd na kunstmatige infectie van groene internodien op levende takken. Het was zelfs niet mogelijk na verwonding van de bast infectie teverkrijge n van het bastweefsel, waarin nog geenfellogee n was gevormd. Het lijkt daarom waarschijnlijk dat ongunstige groeiomstandigheden de primaire oorzaak zijn van instervings-symptomen, waargenomen bij koffie in Kenya.

References

Bock, K. R., 1956. Investigations on the coffee berry disease. Laboratory studies. E. Afr. agric. J. 22:97-103 . Bock, K. R., 1963. The control of coffee berry disease in Kenya. Emp. J. exp. Agric. 31: 97-107. Butler, E. J., 1918. Fungi and diseases in plants. Calcutta. Butt, D.J . &Butters ,B . 1966. Thecontro l ofcoffe e berrydiseas ei nUganda .Proc . 1st.Spec .Meetin g on Coffee Research, E. Afr. Common Serv. Org., Nairobi, Kenya.

42 Gibbs, J. N., 1969. Inoculum sources for coffee berry disease. Ann. appl. Biol. 64: 515-522. Gray, P., 1954. The Microtomist's formulary and guide. Blackiston Co. Inc. Griffiths, E. & Gibbs, J. N., 1969. Early season sprays for the control of coffee berry disease. Ann. appl. Biol. 64: 523-532. Gutierrez, L. H. de, 1954. Muerte descendente causada por Colktotrichum en las plantas de Cafe en el almacigo y su combate por medio de aspersion en Turrialba, Costa-Rica. Turrialba 4: 115-123. Hindorf, H., 1970. Colktotrichum spp. isolated from Coffea arabica L. in Kenya, (in press). Macdonald, J., 1926. A preliminary account of a disease of green coffee berries in Kenya Colony. Trans. Br. mycol. Soc. 11: 145-154. Macdonald, J., 1937. Coffee in Kenya. Diseases of coffee. Dep. Agric, Kenya: 151-164. Meiffren, M., 1957.Le s Maladies du Cafeier en Cote d'lvoire. Centre. Rech. agron. Bingerville: 67-72. Mendes da Ponte, A., 1966. Spraying of Arabica coffee with calcium superphosphates for the control of coffee berry disease normally attributed to Colktotrichum coffeanum Noack. Kenya Coff. 31: 21-22. Mulder, D. & Hocking, D., 1967. Hypothesis to explain the uneven distribution of coffee berry disease in areas of endemic occurrence. Meded. Rijksfac. LandbWet. Gent 32: 729-734. Muller, R. A., 1964. L'anthracnose des baies du cafeier d'Arabie {Coffea arabica), due a Colktotri­ chum coffeanum Noack, au Cameroun. I.F.C.C. Bull. no. 6. Nicholls, W., 1969. The progress of bark formation in Arabica Coffee. Kenya Coff. 34: 429-434. Noack, D., 1901. Die Krankheiten des Kaffeebaumes in Brasilien. III. Colktotrichum coffeanum n.sp. Z. Pfl.Krankh. PflPath. PflSchutz 11: 202. Nutman, F. J. & Roberts, F. M. 1960. Investigations on a disease of Coffea arabica caused by a form of Colktotrichum coffeanum Noack. I. Some factors affecting infection by the pathogen. Trans. Br. mycol. Soc. 43: 489-505. Nutman, F. J. & Roberts, F. M., 1961.Investigation s on a disease of Coffea arabica caused by a form of Colktotrichum coffeanum Noack. III. The relation between infection of bearing wood and disease incidence. Trans. Br. mycol. Soc. 44: 511-521. Nutman, F. J. & Roberts, F. M. 1969. Seasonal variations in the sporulating capacity of the fungus causing coffee berry disease. Ann. appl. Biol. 64: 85-99. Rayner, R. W., 1948. Latent infection in Coffea arabica L. Nature, Lond. 161: 245-246. Rayner, R. W., 1952. Coffee berry disease - A survey of investigations carried out up to 1950. E. Afr. agric. J. 17: 130-158. Saccas, A. M. & Charpentier, J. 1969. L'anthracnose des cafeiers Robusta et Excelsa, due a Colkto­ trichum coffeanum Noack, en Republique Centra-africaine. I.F.C.C. Bull. no. 9. Small, W., 1926. On the occurrences of a species of Colktotrichum. Trans. Br. mycol. Soc, 11: 122- 137. Thorold, C. A., 1945. Elgon die-back disease of Coffee. E. Afr. agric. J. 10: 198-206. Vermeulen, H., 1968. Screening fungicides for the control of coffee berry disease in Kenya. Expl Agric. 4: 255-261. Vermeulen, H., 1970. Coffee berry disease in Kenya. II. The role of Glomerella cingulata in the Colk­ totrichum population, colonizing the bark of Coffea arabica. Neth. J. PI. Path. 76: 285-292. Wallis, J. A. N. & Firman, I. D. 1965. Spraying Arabica coffee for the control of coffee berry disease. Ann. appl. Biol. 55: 139-148.

43 5.2 Coffee berry disease in Kenya. II.Th e role of Glomerella cingulata inth eColletotrichu m population, colonizing thebar k of Coffea arabica

H. VERMEULEN1

East African Agriculture and Forestry Research Organization, P.O.B. 30148, Nairobi, Kenya

Accepted 15Ma y 1970

Abstract On thematurin g bark ofcu t branches of Coffea arabica previously sprayed with copper fungicides perithecia of Glomerella cingulata were easily found after twot o tenday s of incubation. Without fungicides the number ofperitheci a was decidedly lower. On prunings left onth egroun d underth e coffee trees for 6t o 24 weeks theperitheci a could also befound , butth enumber s declined rapidly with time. Perithecia ofG. cingulatacoul d forcibly discharge ascospores under laboratory conditions. Mono- spore cultures obtained bycatchin g ascospores onagar , invariably belonged tothre e Colletotrichum types. It was rarely possible to isolate a Colletotrichum able to infect green coffee berries. Growth- rate, colour of the mycelium, number of conidia produced invitro an dinfectivit y on green coffee berries, however, differed substantially from C.coffeanum, the cause ofcoffe e berry disease. In Kenya no evidence hasbee n obtained that ascospores from perithecia on bark could infect wounded or unwounded greencoffe e berries.Neithe r hasan yinfectio n beenobtaine d withascospore s from perithecia grown invitro. Possibl e explanations for thedifferenc e with previousfindings ar e offered. Based onth edat a presented inthi s paper, iti sconclude d that G. cingulata isno tlikel yt o play a role inth e epidemiology ofth e coffee berry disease in Kenya.

Introduction

Macdonald (1926) and Small (1926) described the incidence of Glomerella cingulata (Stonem.) Spaud. & V. Schrenk in relation to certain saprophytic isolates of Colleto­ trichumcoffeanum Noac k obtained from Coffeaarabica L .i nEas tAfrica . InWes tAfri ­ ca, Meiffren (1957) and Boisson (1960) published data on the occurrence of G.cingu­ lata also on Arabica coffee. No information on the incidence of G. cingulata under field conditions was available from East Africa until Nutman and Roberts reported at the First Coffee Specialist Conference at Nairobi, Kenya in 1966tha t they had found perithecia of this ascomycete on cut branches of coffee. At that time it was still uncertain which role G.cingulata playe d inth eepidemiolog y of thecoffe e berry disease (CBD), the most serious coffee problem in East Africa, caused by a parasitic Colle­ totrichumspecies , now specifically named C.coffeanum Noac k (Hindorf, 1970). Hocking et al. (1967) reported on findings in Tanzania with G.cingulata, while Nutman and Roberts (1969) provided information on the effects of some fungicides on G.cingulata i n relation to the control of CBD. No further details were, however,

1 Present address: Internationaal Agrarisch Centrum (IAC), Wageningen, the Netherlands.

44 available on the role ot (J.cingulata in the Colletotrichumpopulatio ncolonizing , coffee bark. Thispape r reports on thefollowin g aspects ofG. cingulata : a. the occurrence on sprayed and unsprayed coffee trees, b. thedischarg ean d thedistanc e ofdischarg e of ascospores; c. the various forms of Colletotrichum, which could beobtaine d after passing through theG. cingulata stage ; d. thepathogenicit y ofth e Colletotrichumisolates ; e. thepathogenicit y ofth eascospore s ofG. cingulata.

Materialsan dmethod s

Unsprayed branches from a farm near Nairobi, at 1800 m, or those sprayed with copper from the National Agriculture Laboratories, Nairobi, at 1700 m of 'S.L. 34', 'French Mission', 'Harar' or 'Mocha' varieties of C.arabicawer e cut off; the distal portion of each branch was discarded, the cut being made one internode above the one showing obvious signso fbar k maturation. Thebasa lportion swer eals o discarded, leaving an eight-internode long twig (Nutman and Roberts, 1961, 1969). These were divided serially in screw-topped jars. The first bottle contained the internodes 1+ 2, the second 3 + 4, the third 5+ 6, and the fourth 7+ 8. The internodes were washed either with water containing 10pp m 'Teepol' and afterwards washed in three changes of water or with 0.1% HgCl2 for three minutes and afterwards rinsed in threechange s of water. Prunings left on the ground under coffee trees for periods ranging from 6 to 24 weeks were also collected and treated in almost the same way, although it was very difficult to distinguish any previous signs of the progress of bark maturation on the then moribund tissue. Therefore the three distal internodes were usually discarded and the remaining part of the branch divided serially as described above. The identification of perithecia of G.cingulata was relatively easy as the translucent rostra of the ripe perithecia were typical of this fungus (Fig. 1). None of the other fungi colonizing the bark produced fruiting bodies similar to those of G.cingulata. The discharge of ascospores from perithecia on the bark of cut branches was observed by putting small pieces of the bark material, on which perithecia could be found after two to ten days, facing upwards in the lids of inverted Petri dishes. Verti­ cally discharged ascospores could be found on the agar inverted above the bark material. The horizontal discharge was observed by keeping the Petri dishes with the bark stuck to the lids stored sideways. The distance between bark material and agar could be changed by using different amounts of agar in the Petri dishes. During all experiments the temperatures varied between 22°-24°C. When the ascospores were found on the agar above the bark material, mono-spore isolations were made under the low-power magnification of the microscopeb ycuttin g with a fine, sterile needle small blocks of agar containing one ascospore and trans­ ferring these blocks to malt agar plates. The isolates grown from these transfers were after ten to fifteen daysclassifie d according to Hindorf's system (1970): \.C. coffeanum (Gibbs, 1969: C.coffeanum lvar. virulans'), 2. C.acutatum Simmonds (Gibbs, 1969: ccp). 3. C.gloeosporioidesPenzy (Gibbs, 1969:ccm), 4. C.gloeosporioides, 5. C.gloeo- sporioides (Gibbs, 1969: ccd) and 6. G.cingulata. This classification and nomen-

45 Fig. 1. Section through a perithecium of G. cin- gulata,growin g on the bark of C. arabica. Fig. 1. Doorsnede van een peritheciumvan G. cingulala opde bast van C. arabica.

clature have been used throughout this paper. In case of cultural resemblance with C.coffeanum, infection tests on green coffee berries (cv. 'S.L. 34') were carried out to establish the pathogenicity of the isolate (Bock, 1956). To establish the pathogenicity of ascospores, mature perithecia on bark, previously treated for three minutes with 0.1% HgCl 2 and rinsed with sterile water were selected under the binocular microscope. The contents of the fruiting bodies, showing clearly the translucent rostra of the ripe C.c/«gw/ata-perithecia , were transferred with a fine needle to a bottle with water. After transferring the contents of 100-150 perithecia in this way,th e suspension wascentrifuged . The supernatant water wasthe n decanted, so that a pellet of ascospores and asci remained in the centrifuge tube. This pellet was suspended in one or two mlwate r and theconcentratio n assessed on ahaemocyto - meter. Green berries were then inoculated with this suspension; the berries were divided into wounded (needle-prick after inoculation in the centre of the inoculation- droplet) and unwounded berries. Ripe perithecia of G.cingulata, grown in vitro,wer e treated in the same way. For allthes e tests the berries of the CBD-susceptible 'S.L.-34' variety were used, while also in all tests concurrently a batch of berries was inoculated with a fresh conidial suspension of C.coffeanum of the same density of inoculum as the ascospore suspension. Tests were also carried out to assess the effect of sterile tap-water on the germina­ tion of ascospores. The Petri dishes were partly filled with sterile tap-water. Pieces of coffee bark with ripe perithecia were stuck to the lids of those dishes. Discharged ascospores were observed floating in the water.

46 Results

After four to seven days many ripe perithecia were found on the maturing bark of cut branches of coffee, previously sprayed with copper fungicides. Sometimes mature perithecia occurred even after two to three days incubation, only the numbers were much lower than after four to seven days. It was alsopossibl e to find new ripe perithe­ cia after eight, nine or ten days. Then numbers of maturing fruiting bodies decreased rapidly. The number of perithecia on coffee branches, which had never been sprayed before with any fungicide was 60-90% lower than on sprayed branches. It was still possible, however, to find fruiting bodies of G.cingulata on unsprayed branches. A few perithecia were observed on prunings left for 6 to 24 weeks, but it became increasingly difficult to detect perithecia in the course of this period because of the growth of secondary fungi. Selected at the proper time, the perithecia in the inverted Petri dishes, started to discharge ascospores after approximately sixteen hours at 22°-24°C. About 72 hours after the bark material had been put in the Petri dishes, discharge was no longer observed. Discharge of ascospores also occurred after surface sterilization, although the numbers of spores caught were 20-40% lower than in the case of bark material only washed with water. The ascospores caught on the agar above the bark material were clearly distinguisable (Fig.2) . When water agar was used the ascospores very often produced not only a germination tube and an appressorium. but also a coni- diophore bearing a Colletotrichum conidium (Fig.3) . This was also observedwhe nth e ascospores germinated in sterile tap-water. Observations were made on the distance covered by discharged ascospores both vertically and horizontally. Aclassificatio n was made on how thisdischarg e occurred: the number of spores per discharged projectile in relation to the distance covered (Fig.4) . The data presented in Fig.4 appl y to the vertical discharge range from 0.1-1.2 cm. The horizontal discharge pattern is very similar to that of vertical discharge and the data are therefore not presented in Fig.4 . Projectiles containing four to eight

lo^m

OAJS /

Fig. 2. A group of ascospores of G. cingulata Fig. 3. Single ascospore of G. cingulata caught caught on agar as one projectile. Germination on agar. Germination tube, appressorium, and tubes visible in the agar. conidiophore bearing conidium are clearly dis­ tinguishable. Fig. 2. Een groep van ascosporen van G. cingu­ lata gevangen op agar als een projectiel. Kiem- Fig. 3. Enkele ascosporen van G. cingulata ge­ buizen zichtbaar in de agar. vangen op agar. Kiembuis, appressorium enconi- diophoor met een conidium zijn duidelijk zicht­ baar.

47 Fig. 4. Diagram of vertically discharged Number of projectiles ascospores of G. cingulata in discharge 600' i>2028 distance and number of spores per dis­ charged projectile. Fig. 4. Grafiek van verticaal uitgeschoten 500- ascosporen van G. cingulata in relatie tot o o 0.1 - 0.4 cm afstand van uitschieten en het aantal sporen distance per projectiel. • -a 0.4-0.8c m • vertical discharge 400' 1 i 0.8 -1.2c m

1 \ 300 1 I \ \\ \ \\ \ \ \ 200 \ I \

100- M V A \ /

1 2 3 4 5 6 7 8 Spores per projectile spores were occasionally found at distances of 1.5 cm atmos t for the vertical discharge and 2.0 cm for the horizontal discharge. These observations are also not given in Fig.4 becaus e of the very low numbers. Mono-spore isolates from perithecia previous­ ly sprayed with copper fungicides yielded in about 90% of the transfers isolate 3, in about 5% isolate 5 and 2-3% isolate 6, the saprophytes classified by Hindorf (1970) as C.gloeosporioides (3 and 5) and G.cingulata (6) . In four cases -during the rain periods of April-May 1967, March 1968, October 1968 and March-April 1969- an isolate mildly parasitic on green berries was isolated which was classified as C.gloeosporioides, isolate 4 (Hindorf, 1970). This type was always obtained from material of internodes 1 + 2 and 3 + 4, the younger internodes where the bark matu­ ration had just started. The growth-rate of the greenish mycelium exceeded that of C.coffeaniim, whil e relatively few conidia were produced in vitro. The pathogenicity of the conidia of isolate 4 on green berries at approximately 103con./m l was about 1/4 of the pathogenicity of C.coffeanum at the same density of inoculum. As there were much less perithecia on non-sprayed bark the findings on the distribution of the four types are less reliable. Inoculations with ascospore suspensions, derived from perithecia on bark or perithecia grown in vitro, at concentrations ranging from 104-106 spores/ml. on green coffee berries did not result in obvious lesions on the berries, even after wounding. The C.coffeanum inoculation s carried out concurrently at the same concentration range gave infections varying from 50-90% .

48 Discussion

Kaiser and Lukezic (1966) described the discharge of ascospores of G.cingulata on bananas and Hocking et al. reported the violent discharge of ascospores of the same fungus on coffee in Tanzania. Nutman and Roberts (1969) in Kenya, however, could not catch any airborne ascospores when they left coffee branches carrying large num­ bers of ripe perithecia suspended vertically above a spore trap ina polythene-covere d chamber equipped with a humidifier. Therefore they concluded that G.cingulata on coffee lacked the ability to liberate ascospores directly into the air. The experiments described in this paper and carried out from March 1967-July 1969 show that the ascospores of G.cingulata are discharged forcibly into the air under the conditions provided by the inverted Petri dish technique in the laboratory. Discharge also occurs after surface sterilization of the bark containing perithecia. This eliminates any possibility of contamination by Colletotrichum conidia sticking to the outside of perithecial tissues. The ascospores caught on the agar above the bark material had the typical form of ascospores of G.cingulata. When an ascospore germinated on water agar or in water, it very often produced a conidiophore bearing a typical Colletotrichumconidium . The mono-ascospore isolates always gave rise to Colletotrichumcultures . It is difficult to offer an explanation for the discrepancy between the conclusion of Nutman and Roberts (1969) that ascospores cannot be discharged violently and the findings presented in this paper. It may well be that the difference in the findings are only caused by differences in laboratory techniques. Kaiser and Lukezic (1966) found that moisture was the most important environmental factor affecting ascospore discharge. Discharge only occurred when the perithecia were wet or had been kept at 100% R.H . forfifteen hours .Thi s agreeswit h the findings described in thispape r when discharge started after about sixteen hours in the inverted Petri dishes. The detailed observations by Nutman and Roberts (1969) on the marked effect of spray treatments on perithecial production of G.cingulata could be confirmed. The unsprayed branches in the experiments described in this paper showed, however, a decidedly lower production of perithecia than the control treatments of Nutman and Roberts. This is probably due to the fact that the control material used by Nutman and Roberts may have had a spray history, while the unsprayed material used here had never been sprayed with fungicides in the past. This observation supports the suggestion made by Nutman and Roberts (1969) that fungicidal sprays can influence the relative abundance of certain components of the microflora of the maturing bark and in particular the Colletotrichumspp . with a G.cingulata-stage. Only types 3,4 , 5, and 6 could be grown from mono-ascospore isolates of G.cingulata an d this agrees with the findings of Hindorf (1970). Isolate 3 being the most common Colletotrichum component in the maturing bark at.thi s altitude, it ispossibl e to explain its high level of incidence (Vermeulen, 1970), while the ability of this type to produce perithecia after a relatively short time (Hindorf, 1970) should also be taken into account. C.coffeanum and C.acutatum,wer e never obtained after passage through G.cingulata. The possibility that C.coffeanum would be a danger to the Kenya coffee industry throughth etranspor t of ascospores by air seems therefore to be negligible. The mildly parasitic nature of isolate 4, its low level of occurrence and the apparent low produc­ tivity of conidia in vitro do not seem to indicate that this particular strain will be a

49 threat to a country like Kenya with an already high level of CBD. This type might be of interest in countries with little or no CBD. Although isolate 4 occurred during the same time of the year and in the same internodes as C.coffeanum, this isolate could not be obtained directly from bark (Vermeulen, 1970). This would suggest an even lower level of incidence of isolate 4 than C.coffeanum. Researc h should be continued. Theresult sdescribe d abovefo r theyield s of the Colletotrichumisolate s (approxima­ tely 98% saprophytic), obtained through G.cingulata, agree with those of direct inoculations of green berries with suspensions of ascospores from perithecia either produced on branches or invitro. Eve n after wounding no infections could be induced. The ascospore germination was not affected by the tap-water used, as the spores germinated easily in sterile tap-water. These results do not agree with the previous findings of Hocking, Johanns and Vermeulen (1967) who reported successful infec­ tions on berries, using ascospores from perithecia of G.cingulata on green coffee berries. Several possible explanations for the difference in findings may be offered: 1. The inoculation experiments reported by Hocking et al. from Tanzania were carried out with ascospores from perithecia on green berries and not from perithe­ cia on branches or perithecia grown in vitro. 2. C.coffeanum, introduce d only recently in the main coffee area of Tanzania, may not yet have lost its capacity to produce perithecia in that country1. 3. C.gloeosporioides, isolate 4 (Hindorf, 1970) is more prevalent in Tanzania and has been mistaken for C.coffeanum, becaus e of its capacity to cause infection on green berries. 4. The level of saprophytic components of the Colletotrichum-population is so high in Kenya, that the rare perfect stage of C.coffeanum, becaus e of the low incidence of the parasite, has been overlooked. The observations made on the discharge of ascospores in relation to discharge distance agree with the findings of Ingold (1965, p.52-74) . The discharge distance is amply sufficient to make the spores airborne. As it was hardly likely that G. cingulata plays a role in the epidemiology of C.coffeanum i n Kenya, no further investigations were carried out to assess the liberation of ascospores under field conditions.

Acknowledgments

I am grateful to the Ministry of Agriculture of Kenya for the research facilities and help during February 1967 - September 1969. This paper is published with kind permission of the Director of the East African Agriculture and Forestry Research Organization, Nairobi, Kenya and the Ministry of Foreign Affairs of the Netherlands.

.Samenvattin g

Koffiebesziekte in Kenia. II. De rol van Glomerella cingulata in de Colletotrichum populatie, diede bast vanCoffea arabica koloniseert.

Op afgesneden takken van met fungiciden bespoten Coffea arabicakonde n de peri- thecienva nGlomerella cingulata (Fig .1 )i n grote hoeveelheden worden gevonden op de 1 J. C. Hudson has found recently again perithecia of G. cingulata on CBD-affected berries in Tan­ zania (pers.coram.) .

50 rijpende bast na twee tot tien dagen incubatie. Op takken, die nog nooit met fungici- den waren bespoten, lagen de aantallen perithecien aanzienlijk lager. Ook werd G.cin- gulata gevonden op snoeihout onder de koffiebomen, maar de aantallen perithecien verminderden snel na het tijdstip van snoeien. Het bleek mogelijk te zijn om onder laboratoriumomstandigheden perithecien van G.cingulata ascosporen in de lucht te laten schieten (Fig.2) . Ook nadat de takken uitwendig waren ontsmet met sublimaat werden ascoporen uitgestoten. De opge- vangen sporen vormden vaak op water agar en in water niet alleen een appressorium en een kiembuis, maar soms ook een conidiophoor met een Colletotrichumconidiu m (Fig.3) . De maximale afstand afgelegd door uitgeschoten ascosporen bedroeg verti- caal ongeveer 1.5 cm, horizontaal ongeveer 2.0 cm (Fig.4) . Een monosporenisolatie van een uitgeschoten en op agar opgevangen ascospore gaf altijd eenva n vier typen die tot de Colletotrichum populatie behoren. Een van deze vier bleek een C.gloeosporioides te zijn, die in staat was groene koffiebessen te infec- teren. Groeisnelheid, kleur van het mycelium, het aantal gevormde conidien in vitro en de pathogeniteit op groene koffiebessen bleken echter aanzienlijk te verschillen van C.coffeanum, de veroorzaker van de koffiebesziekte. Er konden geen bewijzen worden gevonden, dat ascosporen afkomstig van perithecien op de koffiebast in staat waren gewonde of niet-gewonde groene koffiebessen te infecteren. Ook ascosporen afkomstig van perithecien van G.cingulata in vitro konden geen groene koffiebessen aantasten. Verschillende mogelijkheden werden gesuggereerd om het verschil te verklaren tussen deze waarnemingen en eerder gepubliceerde gegevens, waaruit bleek dat ascosporesuspensies verkregen van perithecien op groene bessen weli n staat waren geweest infectie op koffiebessen te veroorzaken.

References

Bock, K. R., 1956. Investigations on coffee berry disease. Laboratory studies. E. Afr. agric. J. 22 97-103. Boisson, C, 1960. L'anthracnose du Cafeier. Revue Mycol. 25: 263-292. Gibbs, J. N., 1969. Inoculum sources for coffee berry disease. Ann. appl. Biol. 64: 515-522. Hindorf, H., 1970. Colletotrichum spp. isolated from Coffee arabica L. in Kenya (in press). Hocking, D., Johanns, J. C. & Vermeulen, H., 1967. Ascospore production, discharge and infection by Glomerella cingulata causing coffee berry disease. Nature, Lond. 214: 1144-1145. Ingold, C. T., 1965. Spore Liberation. Clarendon Press, Oxford. Kaiser, W. J. & Lukezic, F. L., 1966. Occurrence, sporulation and pathogenicity studies with Glome­ rella cingulata associated with Crown rot of boxed bananas. Mycologia, 58: 397-405. Macdonald, J., 1926. A preliminaryaccount of a disease of green coffee berries in Kenya. Colony. Trans, mycol. Soc. 11: 145-154. Meiffren, M., 1957. Les maladies du Cafeier en Cote d'lvoire. Centre Rech. agron. Bingerville: 67-72. Nutman, F. J. & Roberts, F. M., 1961. Investigations on a disease of Coffea arabica caused by a form of Colletotrichum coffeanum Noack. III. The relation between infection of bearing wood and disease incidence. Trans. Br. mycol. Soc. 44: 511-521. Nutman, F. J. & Roberts, F. M., 1969. The stimulating effect of some fungicides on Glomerella cingulata in relation to the control of coffee berry disease. Ann. appl. Biol. 64: 335-344. Small, W., 1926.O n the occurrences ofa species of Co/toofnc/uw!. Trans. Br. mycol. Soc. 11:112-137. Vermeulen, H., 1970. Coffee berry disease in Kenya. I. Colletotrichum spp. colonizing the bark of Coffea arabica L. Neth. J. PI. Path. 76: 277-284.

51 5.3 Ascospore Production,Discharg ean dInfectio nb y Glomerella aingulata causing CoffeeBerr y Disease (reprinted from Nature,, Vol. 214, No 5093, pp 1144-1145, June 10, 1967) Theanthracnos eo fgree ncoffe eberrie sknow ni nEas tAfric aa s 'coffeeberr y disease' (CBD)cause ssever elosse si nAfrica ncountrie swhic hgro wcoffee .I n Kenyath eeconomi c loss lastseaso nwa sestimate da t£ 3,000,00 0o rsom e2 0pe r cento fth eoveral l crop.T othi smus tb eadde d another£ 250,000 ,th ecos to f controlmeasures 1. Thereha sbee nconsiderabl eresearc ho nth ediseas ean dit scausa l fungus,a virulent formo f Colletotriahum aoffeanum Noack2-8 (aconidia lstat eo f Glomerella aingulata (Stonem.)Spauld .an dSchenk) ,a swel la so nth euniversall y occurring 'brownblight' ,a relate ddiseas erestricte di ncoffe et orip ecoffe eberrie san d causedb yles svirulen t formso fth esam e fungus9'10.Th evirulen t formi sdis ­ tinguished fromothe rform sonl yb yit sabilit yt oinfec tgree nberrie san dcanno t bedistinguishe do na consisten tmorphologica lbasis 2'11. CBDi spredominantl ydisperse d locallyb yrai nsplas ho fconidi aproduce db y saprophytic infectionsi nth ebar ko fth ebearin g twigs12'13. Dispersalove rlon g distancesca nan ddoe soccu rthroug hth emovemen to finfecte d seedlings.Thi si s shownb yth epresenc ei nnon-endemi c areaso fisolate d treeso rsmal lgroup so f treeso nwhic h everybranc hbear s infectedberries .Thi sdoe snot ,however ,ex ­ plainth eoccurrenc eo fne woutbreak st owhic hn omovemen to fseedling sca nb e traced,an di nwhic honl ya ver yfe winfecte dberrie sar efoun do nisolate d single brancheso ntree sotherwis e free fromdisease .Lon gdistanc e transmissiono fdis ­ easeb yconidia ldispersa l throughhuma no rothe ragencie si sunlikel ybecaus eo f thesusceptibilit yo fconidi at odesiccation 1'*.Th eobviou spropagul efo r in­ vestigationsi sth eascospore ,whic hplay sa nimportan tpar ti nth etransmissio n ofman yothe rplan tdiseases . Peritheciawit hascospore so fbrow nbligh t formso f G. aingulata havebee n foundo ntwigs 9'15an do nleaves 16 ofarabic acoffee ,bu tearlie rworker shav e failedt ofin dth eperfec t stateo fth efor mcausin gCBD 2'3'5-8. Peritheciahav e beenfoun di naga rculture s saidt ob eo fth evirulen t form6'8'16'17,bu tn otest s ofpathogenicit y eithero fth ecultur eo ro fth eascospore shav ebee nreported . Duringou rinoculatio nexperiment swit h isolateso fth evirulen t formfro ma numbero fsite si nKenya ,fertil eperitheci awer eforme do noccasiona l green berriesafte rprolonge d incubation.Afte r thisfinding ,larg enumber so finfecte d greenberrie swer ecollecte di nth efiel d fromvariou s sitesan dincubate di n moisturechamber sfo r5 week sa tambien t laboratoryconditions .1 3pe rcen to f thembor eabundan t fertileperithecia .Furthe r searchesi narea si nwhic hCB Di s

52 endemic revealeda noccasiona l berry,sti lo nth etree ,bearin gperithecia .A totalo feigh t suchberrie swa sfound . Atth esam e timew elooke dfo rperitheci ao nothe r tissues.Twi gpriming san d fallen leaveswer e incubatedan dinvariabl y boreperitheci ai nth elaboratory . Pruningswit habundan tperitheci awer e found occasionally inth efield ,particu ­ larlyi nwe tseason san dareas .Th edimension so fasc ian dascospore s fromal l sourcesar eshow ni nTabl e 1.

Table 1:Dimension so fasc ian dascospore s fromal lsources .

Mean Range

Asci 61x 10 u 53-76x 8-12 u Ascospores 16x 4u 12-23x 3 -5 u

The ascosporeswer eunicellular ,slightl y curvedan dslightl ypointe d towards theends ,wit ha distinc t central refractive area;the y conformedi never ywa yt o Glomevella cingulata18. Ascospores from twigs,leave san dberrie s germinated readily in vitro oro n thesurface so fcoffe eberries ,becomin g generally one-septatean dformin gap - pressoria.Culture so naga ro fsingl eascospore s fromtwig swer e infectivet o greenberrie si n1 5pe rcen to ftrials .Culture so fsingl eascospore s from green berries invariably produced infective conidiawhic hi nth efirs t generationwer e 80pe rcen ta svirulen ta sth esam edensit yo finoculu m collected from lesionso n berries inth efield . By repeatedlywashin g largenumber so fperitheci a fromgree nberrie s until thewashing swer e free fromconidia ,pur e suspensionso fascospore s (10 spore/ml) wereprepared .Thes ewer edirectl y infectivet odetache dunwounde d greenberrie s (7pe rcent )an dt odetache dunwounde d internodalregion so fgree ntwig s (5pe r cent).Th elatte rhav epreviousl ybee n foundt oresis tdirec t invasionb yconidia l inoculation6. Byplacin gmatur eperitheci ao nth elid so finverte dPetr idishe san dfindin g ascospores stuckt oglas s slides suspended above,an db yobservin gth egrowt ho f cultureso naga r invertedabove ,i twa sshow ntha tth eascospore swer e discharged violently.Th emaximu mheigh to fvertica lprojectio nwa s2. 2cm .Thi s agreeswit h theresult sfo r G. eingulata fromothe rhosts 19 andi sampl efo rsuc h sporest o become airborne. Greenberrie s inoculatedwit hascospore so rwit h conidia frommonoascospor e culturesusuall ybu tno tinvariabl y gave riset ofurthe rperithecia .Thi s supports the contention thatcertai n strainshav ea ninheren t capacityt ofor mperithecia .

53 The greater frequencyo fperitheci ao ntwig san dleave s thano nberrie sma yb e explainedb yth eobservatio n thatsuc htissue sar emor e likelyt ob einvade db ya greatergenetica lvariet yo fstrains .Considerin gth ecomplexit yo fth egenetic s of sexualityi n G. cingulata2", iti sno tsurprisin g that isolated strainso n berriesar eperithecia l only infrequently. Althoughi tremain st ob esee nwhethe ro rno tthes eascospore sar emor ere ­ sistantt oadvers e conditions thanar econidia ,thi si susuall yth ecas ewit h similar fungi,on eo fth eprincipa l functionso fth esexua l stagebein gsurvival . Thusou rdiscover yo fth eoccurrenc ean dinfectivit yo fascospores ,particularl y togree ntwigs ,implie s thatthes ema yb ea nimportan t sourceo fne woutbreak so f coffeeberr ydisease . A second important functiono fsexua l reproductioni nfung ii sgeneti csegre ­ gationan drecombination .Th eabilit yt oinfec tgree ncoffe eberries ,however ,ap ­ pearst ob econtrolle db ya singl egen e locusan di sno ta quantitativ e factor1*. Thediscover yo fth eoperatio no fsexua lreproductio ni nth efor mo f G. cingulata which causedCB Ddoe sno ttherefor epromot eth epossibilit yo fa new ,ye tmor e virulent,for marisin g throughrecombination .O nth eothe rhand ,i tdoe sno tex ­ cludeth eoccurrenc eo ffurthe rmutation ,th eprobabl eorigi no fth epresen tviru ­ lentform . We thankth eDirecto ro fth eCoffe eResearc h Station,Kenya ,fo rprovisio n offacilities ,an dD r D.flulde rfo rstimulatin g some lineso finvestigation .

DRAKEHOCKING * JOHANNAC .JOHANN S EastAfrica nTropica l PesticidesResearc h Institute,Arusha ,Tanzania .

HANSVERMEULEN t CoffeeResearc h Station,Ruiru ,Kenya .

ReceivedJanuar y 13, 1967. * Present address: 11511-75Avenue ,Edmonton ,Alberta ,Canada . | Present address:Nationa lAgricultura l Laboratories,Nairobi ,Kenya .

1Kenya Coffee, 31,45 0(1966 ) 2Macdonald,J. , Trans.Brit. Mycol. Soa. 11,14 5(1926 ) 3Rayner,E.W. ,E. Afric. Agric. J., 17,13 0(1952 ) ''Bock,K.R. , E. Afric. Agric. J., 22,9 7(1956 ) 5Nutman,F.J .an dRoberts ,F.M . Trans. Brit. Myaol. Soc, 43,48 9(1960 ) sBoisson,C , Rev. Mycol., December issue,26 3(1960 ) 7Muller,R.. ,Bull. Inst. Franc. Cafe Cacao No. 6, 38(1964 )

54 8Saccas,A.M. ,workin gpart yo ncoffe eproductio nan dprotection ,Brazil , Document de Travail Ce/b^/Tti (1965) 9Small,W. , Trans. Brit. Myaol. Soc, 11,11 2(1926 ) 1"Hocking,D. , Ann. App. Biol., 58,40 9(1966 ) 1locking ,D. , Trans. Brit. Myaol. Soc, (inth epress ) 12Nutman,F.J .an dRoberts ,F.M. , Trans. Brit. Myaol. Soc., 44,51 1(1961 ) 13Bock,K.R. , Emp. J. Exp. Agric., 31,8 8(1963 ) 11(Nutman,F.J .an dRoberts ,F.M. , Trans. Brit. Myaol. Soc., 43,64 3(1960 ) blocking1 ,D. , Pans B., 11,41 0(1965 ) lsAnn. Rep. 1962/63, Coffee Res. Station, Ruiru and Coffee Res. Services, Kenya, 54(1963 ) 17Hendrickz,F.L. , Publ. Inst. Nat. Etudes Agric. Congo Beige, Serie Sai., 26, 10(1963 ) 18VonSchrenk ,H. ,an dSpaulding ,P. , U.S. Dept. Agric. Bull. No. 44,5 1(1903 ) 19Kaiser,W.J. ,an dLukezic ,F.L. , Myaologia, 58,39 7(1966 ) 20Wheeler,H.E. ,an dMcGahen ,J.W. , Amer. J. Bot. , 39,11 0(1952 )

Printedi nGrea tBritai nb yFisher ,Knigh t& Co. ,Ltd. ,St .Albans .

5.4Conidia l germinationan dinfectio n

5.4.1 Introduction

NUTMANan dROBERT S(1960a ,b ,1961 ,1969a )assume dtha tth eCB Dpathoge nha dcom ­ pletelyreplace dth esaprophyti c Colletotriahum spp.i nth ematurin gbar ko fcoffee . Wheni twa sfound ,however ,tha tth ebar ki scolonize db ya numbe ro f Colle­ totriahum spp.(GIBBS ,1969 ;HINDORF ,1970 ,1972 ,1973a ;VERMEULEN ,1970a ,b )i t becameapparen ttha tth econidi aproduce do nth ebar ki nth epas tusuall yreferre d toa s'inoculu mpotential '(NUTMA N& ROBERTS ,1961) ,woul db ea mixtur eo fconidi a ofth e Colleiotrichum spp.,inhabitin gbar ktissues . Experimentswer ecarrie dou tt oinvestigat ewhethe ro rno tth econidi ao f thedifferen t Colletotriahum specieswoul daffec teac hothe rdirectl yo rindirectl y duringgermination ,thu sinfluencin gberr yinfectio nb y C. aoffeanum. Thesein ­ cludedth efollowin gaspects :a )th egerminatio no fconidia ,eithe ri na suspensio n froma singl eisolat eo ra conidia lmixtur eo f Colletotriahum spp.i nwater ,o n agaran do nberr ysurfaces ;b )th eeffect so fconidia lexudate so fsaprophyti c Colletotriahum spp.o nth egerminatio no fan dsubsequen tinfectio nb y C. aoffeanum

55 conidiarespectivel yo naga ran dberr ysurfaces ;an dc )th eeffect so nberr y infectiono fconidia lmixture so fsaprophyti c Colletotriohum spp.an d C. ooffeanum. InTabl e8 th ecalculate dnumber so f C. ooffeanum conidiapresen ti nth e germinationo rinoculatio ndroplet s (0.1an d0.02 5m lrespectively) ,a tvariou s 7 concentrationlevel s(10-1(10 -1 00 conidia/ml c )an di nmixture swit hsaprophyti c Colletotriohum spp.ar egiven .

Table8 :Calculate dnumber so f C. ooffeanum conidiapresen ti nvariou sdroplet s andmixture swit hsaprophyti c Colletotriohum spp.a ta rang eo fconcen ­ trationlevel s(se e4.5 , 4.6).

Numberso f C ooffeanum conidia in Colletot- Conidia riohum mixtures (pathogen: saproph y te) Numbersc f concen­ Droplet conidiap er tration path./ path./ path./ path./ droplet perm l sapr. sapr. sapr. sapr. 1: 4 2: 3 3: 2 4: 1

10 0. 25 0.05 0.1 0.15 0.2 102 2. 5 0.5 1 1.5 2 103 25 5 10 15 20 4 0.025ml * .o 250 50 100 150 200 105 2.500 500 1.000 1.500 2.000 106 25.000 5.000 10.000 15.000 20.000 7 .o 250.000 50.000 100.000 150.000 200.000

10 1 0.2 0.4 0.6 0.8 102 10 2 4 6 8 103 100 20 40 60 80 0.1 ml** >04 1.000 200 400 600 800 105 10.000 2.000 4.000 6.000 8.000 6 .o 100.000 20.000 40.000 60.000 80.000 107 1.000.000 200.000 400.000 600.000 800.000

* Dropletsuse di ninfectio ntest so ngree nberries . **Droplet suse di ngerminatio ntest so nagar .

5.4.2 Results of germination and infection experiments

Thegerminatio no fth econidi ao fal l Colletotriohum spp.prove dt ob eslightl y lowero nwate raga ran di nhangin gdrop so fsteril ewater ,tha nth egerminatio no n

56 berrysurfaces .O nberrie sth econidi ao fth esaprophyti c Colletotriohwn spp. oftenforme dappressori awithout ,however ,causin ginfectio no fth eberries .Onl y thecombine dresult so fsi xserie so fgerminatio ntest so nwate raga rar epresente d inFig .7 . Thegerminatio nrate so feac ho fth e Colletotriohwn spp.separatel ydi dno t differsignificantl ya tan yo fth econcentration stested .A tver ylo wconcen ­ trationssingl econidi aofte ndi dno tgerminate .Abov eapproximatel y10 5conidi a perm lth egerminatio nrat eo fth econidi ao fal l Colletotrichum spp.decline d slightly.Th egerminatio no fconidi aharveste dfro mpur eculture sappeare dt ob e lowertha nth egerminatio no fth eCB Dfungu sconidi afro mberries . Conidialmixture so fCB Dwit hconidi ao feac ho fth eothe r Colletotrichum spp. didno tyiel dgerminatio nrate so naga ran do nberr ysurface sdifferen tfro mthos e observedseparately .

100-

/ fc^:X

c O a f SW x x CBD (from berries) l_ +— + CBD (from culture)

-i 1 1 1— 0 3 k 5 6 log.conidia/m l

Fig.7 :Conidia lgerminatio no f C. ooffeanum (fromaffecte dberrie san dpur e cultures)an dsaprophyti c Colletotriohwn spp.(fro mpur ecultures )o n wateraga rafte r8-1 0 hoursa t22°C .

57 Tenserie so ftest so naga rslide san dsi xserie so ftest so nberrie swer e carriedou tt oasses sth eeffect so fconidia lexudate so feac ho fth e Colletot­ riohum spp.,separatel yo rmixed .Neithe ro nth eaga rslide sno ro nth egree n berriessignifican teffect scoul db edetecte do fan yo fth eexudate so fsapro ­ phytic Colletotriohum spp.o nth egerminatio no fan dth esubsequen tinfectio nb y C. coffeanum conidia. Theresult so fnin eexperiments ,i nwhic hth einfectio nrat eo ngree nberrie s wasstudie da sa functio no fth econidia lconcentratio no f C. coffeanum andth e ratiobetwee nparasiti can dsaprophyti c Colletotriohum spp.,ar epresente di n Fig.8 an dTabl e9 .The ysho wthat : -th erat eo finfectio nincrease swit hincreasin gconcentration so fconidi au pt o aconcentratio no f1 0 conidiape rml ; -abov ea concentratio no f1 0 conidiape rm lth einfectio nrat eslightl yde ­ creases; -i nmixture so f C. coffeanum andth esaprophyti c Colletotriohum species,th ein ­ fectionrat ei spositivel ycorrelate dwit hth epercentag eo f C. coffeanum inth e mixturesa teac hconidia lconcentratio nlevel ; -unde rn ocondition sa ninfectio nrat ehighe rtha n75 1ha sbee nachieve di nthes e trials.

Table9 :Infectio npercentage so ngree nberrie s (a)o fa rang eo fconidia lmixture s of C. coffeanum withsaprophyti c Colletotriohum spp.i nrelatio nt oth e calculatednumbe ro f C. coffeanum conidiapresen ti na ninoculatio ndrop ­ leto f0.02 5m l(b) .

Percentage of C. coffeanum inconidia lmixture s (0.025ml ) Concen­ tration 20 40 60 80 100 ofconidi a perm l

7 10 12 50.000 25 100.000 40 150.000 55 200.000 70 250.000 106 15 5.000 28 10.000 45 15.000 60 20.000 75 25.000 .o5 20 500 32 1.000 50 1.500 62 2.000 70 2.500 >o4 10 50 15 100 22 150 30 200 32 250 .o3 2 5 4 10 6 15 12 20 15 25

58 100- * -A 80 % sapr-20% CBD o a 60%, sapr.-40°/oCBD • • 40% sapr.-60°/oCBD v v 20% sapr.-80%CBD 100% CBD 0)

XI 50- at •*->o

0 3 4 5 log.conidia/ml

Fig. 8: Infection rateo n green,expandin g berries inoculated with C. ooffeanvm conidia andwit h arang e ofmixture s of conidia of C. ooffeanvm and saprophytic CoVtetotriahum spp.a tdifferen t concentration levels.

5.5Discussio n

Inth etissue so fth eolde r internodes,viz . internodes 9-14,th esaprophyti c Colletotrichum spp.coul db efoun di nconsiderabl e numbers,initiall y increasing withth eadvanc eo fth ebar kmaturation ,bu tdeclinin g againi nth eoldes ttw o internodes (13-14)t oa lo wleve l (VERMEULEN, 1970a).Th epresenc eo f C. ooffeanum was foundt ob erestricte d tothos e internodeswit hth ever yfirs t signso fbar k maturation,i.e .i ncortica l tissueswher e onlyon ephelloge nwa sbein g formedo r had justbee n formed outsideth estone-cel l layer.Thes e internodes stillhav ea smoothsurface ,whic hi so fimportanc ei nrelatio nt oth etenacit yo ffungicide s duringth erain yseaso n (6.6.,7.2) .Consequentl y theparasit e isrestricte dt oa smallbar kare ajus t apicalo fth ezon ewit hadvancin g stageso fbar kmaturation . Asha sbee nshow nb yNICHOLL S (1969)thes earea swit hth efirs tsign so fbar k

59 maturationca nexten dappreciabl yi ngrowt hflushe sdurin gth eearl ypart so fth e rainyseason .Th eleve lo f C. aoffeanum inth e Colletotriohum populationwil l thereforemos t likelyincreas edurin gth erains ,becaus eo fth eaforementione d enlargemento f C. aoffeanum containing tissues.I nsection s6. 3an d6. 4thi sin ­ creasei n C. aoffeanum levelwil lb efurthe rdiscusse di nconjunctio nwit hth e actualpathogenicit yo fth ebar k inoculum. Thefact ,tha t C. aoffeanum cannotoccup ybrow ncortica l tissueswa stenta ­ tivelyattribute db yVERMEULE N (1970a)t oth enutritiona lrequirement so fth e parasite.Althoug hth enutritiona lpreferenc ecanno tb eexclude d completely, otherstil lunknow nfactor sar eprobabl y involved,permittin gth epathoge nno tonl y tocoloniz ehabituall ywell-define d areasi nbar ktissues ,bu tals ot oattac kth e coffeea tth evariou sstages .Th einabilit yo f C. aoffeanum toinvad eyounger , green internodal tissues (internodes1-6 ,Fig .5 )an dthu st oinduc eElgo ndie - backsymptoms ,a swa sallege db yNUTMA Nan dROBERT S (1960a),wa salread ydemon ­ stratedb yTHOROL D (1945),HOCKIN G (1966)an dagai nconfirme db yVERMEULE N (1970a). RecentlyRAMO San dSHAVDI A (1976)showe dtha tElgo ndiebac ki scause db y Pseudomonas syringae v. Hall. The isolationsmad eb yVERMEULE N (1970a)fro mbar kmateria lyielded ,a swa s established later,a compositio no fth e Colletotrickum populationtypica lfo r coffeebranche sa ttha taltitud e (approximately 1700m a.s.l.)an dsubjecte dt oa coppersprayin gregime .HINDOR F (1973c,1974 )an dMULING E(1971b )foun dmarke dvari ­ ationsi nth e Colletotrickum composition,dependin go ndifference s inaltitude . Considerable informationi sno wals oavailabl eo nth eeffect so ffungicide so nth e compositiono fth e Colletotrickum complex,inhabitin gth ecoffe ebar k (FURTADO, 1969,1970 ;GIBBS ,1969 ,1972 ;GRIFFITHS ,1972 ;HINDORF ,1970 ,1972 ;VERMEULEN , 1970a, b). Withthes efinding so ndistributio nan dcompositio no fth e Colletotri- ahum barkpopulation ,th einoculu mpotentia l theory (BOCK,1963 ;NUTMA N& ROBERTS , 1960a,b ,1961 )ha dt ob ereconsidere dan dre-evaluated .Th econsequence so fth ere ­ considerationo fth einoculu mpotentia l theoryfo rdiseas econtro lar ediscusse di n section7.2 . Theproductio no fmixe dpopulation so f Colletotviohum conidiama yhav e-afte r disseminationthroughou tth ecoffe etre e (WALLER,1972a )-a ninfluenc eo nth e infectiono fberries .Therefore ,th egerminatio no fconidi ai nmixe dpopulation san d thesubsequen t infectiono fberrie sb ythes econidi awer estudie dexperimentally . Thegerminatio no f C. aoffeanum conidiao nwate raga rwa sno taffecte db yth e presenceo fothe r Colletotrickum spp.,althoug hbot hi npur ean dmixe dconidia l suspensionsth egerminatio nrat edecline dabov eapproximatel y 10^conidi ape rml . Thisdeclin e-als onotice db yNUTMA Nan dROBERT S (1970)-migh tb eattribute dt oa 'lumpingeffect' ,i.e .conidi aremai nclustere di nconcentrate d suspensionsan d

60 2 henceth egerminatio nrat edeclines .I nver ydilute dsuspension s (10t o1 0 conidiape rml )singl econidi aofte ndi dno tgerminat e forreason s stillunknown . AlsoSTEINE R (1973a)notice dbot hth e 'lumpingeffect 'an dth eabsenc eo fgermi ­ nationo fsingl espores . Noeffect so fth eexudate so fth esaprophyti c Colletotriahum conidiawer eob ­ served inth eserie so fexperiments ,eithe ro nth egerminatio no f C. aoffeanum conidiao naga ro rth epathogenicit yo f C. aoffeanum conidiao nberries .Mor econ ­ clusiveevidenc ewhethe ro rno texudate so f Colletotriahum conidiawoul da tan y timeexer teffect so nth egerminatio no fan dinfectio nb yth econidi ao fCB Di s stillrequire dan dmor eresearc hshoul db ecarrie dou tt oascertai nth eobserva ­ tionspresente d inthi spublication . Theresult so fth elaborator y infectiontest swit h Colletotriahum conidial mixtureso nberrie sals o showlik eth egerminatio nrat ea declin eo fth e infection rateabov eapproximatel y 10 conidiape rm l forbot hpur eCB D suspensionsan dth e mixtures.Thi sdeclin ema yagai nb eattribute d toth e 'lumpingeffect 'a smen ­ tionedbefore .Apar t fromth ephysica lobstructio nexerte db y thelumpe dconidi a oneac hother ,i ti sals opossibl etha tth ecluster so fconidi awil lno treceiv e thegerminatio nimpulse sfro mth eberr y surface,a sreporte db ySTEINE R (1972b) forth eberrie so fCB Dsusceptibl ecoffe ecultivars . Thedeclin ei nth einfectio nrate ,a tconcentration sabov e 10 conidiape rml , is,however ,o facademi c interestonly ,a ssimila rhig hconidia lconcentration sd o notoccu runde r fieldcondition s eveni nth eperiod sconductiv e for infection,i.e . therain yseaso n (WALLER,1971 ,1972) . Forconcentration sbelo w 103 conidiape rml ,a sstate dbefore ,eithe ra thresholdeffec to ra reductio no fth e infectionprobabilit ymigh thav ecause dth e declineo finfection .Fro mTabl e9 i tappear stha ta ta give nconcentratio no fth e conidial suspensionth epercentag e of infectioni sproportiona l toth e increasing numbero fconidi ao f C. aoffeanum, notdependin g onth enumber so fconidi ao fth e saprophytic Colletotriahum componentso fth einoculatio ndroplets.(FURTADO ,1969 ) andGRIFFITH San dFURTAD O (1972)wer eals ounabl et ofin dan ydistinc t antagonistic effectsexerte db y thesaprophyti c Colletotriahum spp.o n C. aoffeanum. Antagonis­ ticeffect smay ,however ,b eexcerte db yothe rmicro-organism spersistin g oncoffe e branches (WALLER, 1972a). Bysprayin g fungicides thetota lsporulatin g capacityo fth e Colletotriahum fungiinhabitin g thebar k isreduced ,a swa sshow nb yBOC K (1963),NuTMA Nan d ROBERTS (1961)an dwa sconfirme db yth efinding spresente d inChapte r 6, viz. thesporulatin gcapacit ydat afro mMeru .GIBB S (1972)establishe dtha tth e

61 sporulating capacityo f C. coffeanwn soonros eafte rth eterminatio no fspra yap ­ plications (inth ecas eo fpre-rai nsprays) .I ncombinatio nwit hth eafore ­ mentioned enlargemento f C. coffeanwn containing tissuesdurin gth erain san d thusth eexpecte dhighe rconidia lproductio no f C. coffeanwn, thiswoul d leadt o ahighe r levelo f C. coffeanwn conidiadurin gth erains .Hence ,i ndroplet swit h conidial suspensionsdeposite do nberries ,a shif t from 20% to80 % CBDa tth e normalconcentratio n levelo f1 0 conidiape rm l (WALLER,1972a )woul d increase the infectionprobabilit y from2 %t o12 % (Table 9), thusenhancin gth eimpor ­ tanceo fbar k inoculumo f C. coffeanwn, availablefo rberr y infection,consider ­ ably. When' G. cingulata wasfoun do ncoppe r sprayed trees (NUTMAN& ROBERTS ,1969c) , mostattentio nwa sthe ngive nt oth erol ethi sperfec tstag ewoul dpla yi nth e epidemiologyo fCBD .NUTMA Nan dROBERT Swer eunabl ei nrepeate d laboratoryexperi ­ mentst ocatc hair-born eascospore so nspor e trapsuse di nthes etrials .However , theresult spresente db yVERMEULE N (1970b)an dreprinte dher e(p .44) , showtha tasco ­ sporesar eejecte d forciblyove rdistance so fabou t1. 2c mvertically ,a distanc e sufficientt omak eth espore s air-borne.Fro msubsequen tnumerou smono-ascospore s isolationsonl ytw oform so fsaprophyti c C. gloeosporioides, amild yparasiti c Colletotvichum sp.an d G. cingulata couldb eobtained .Neithe r C. coffeanwn nor C. acutatuvi wereeve robtaine d from G. cingulata. Despiteman y attempts,ther ear e norecord so fth epresenc eo fth eperfec tstag eo f C. coffeanwn, havingbee nfoun d inKenya .A sascospore so f G. cingulata were foundt ob eincapabl et ocaus ein ­ fectiono ngree nberries ,VERMEULE N (1970b)conclude d that G. cingulata didno t playa par ti nth eepidemiolog yo fCB Di nKenya . InTanzania ,however ,HOCKIN Ge tal . (1967)foun dperitheci ao f G. cingulata ongree ncoffe eberrie s afterinoculatin g thoseberrie swit ha CB Dsuspension . HINDORF (1972)cite danothe r instanceo fth epresenc eo f G. cingulata onberrie s inTanzani a (JOHANNS,persona l communication),givin g riset oa C. gloeosporioides, capablet ocaus ea lo wleve lo fberr y infection.VERMEULE N (1970b)suggeste da numbero fpossibilitie st oexplai nth edifference sbetwee nth eKeny aan dTanzani a findings,e.g .tha t C. gloeospovioid.es hadbee nmistake nfo r C. coffeanum, because ofit scultura lsimilaritie san dtha t C. coffeanwn, newlyreporte di nTanzania , hadno tye tlos tit scapacit yt oproduc eperithecia .FIRMA Nan dWALLE R (1977)con ­ sideredth efirs tsuggestio nunlikel yan dpointe dou ttha t- wit hregar dt oth e latterexplanatio n-th eCB Di nTanzani awa smos t likelyintroduce d fromKenya , whereth efungu swa sno tknow nt oproduc eperithecia . Asth ematte ro fth ediseas eepidemiolog yha sno tye tbee nsettle dan dth e suggestiono ftranspor to fCB Dwit hplantin gmateria l (seedlings),forwarde db y

62 FIRMANan dWALLE R (1977),i sstil lwithou t sufficientresearc hevidence ,th e discrepancy between theresult s obtainedwit h G. oingulata inbot hKeny aan d Tanzania remains unresolved.

63 6. SPORULATION AND PROBABILITY OF INFECTION IN RELATIOr TO RAINFALL AND TIMING OF FUNGICIDE APPLICATIONS

6.1 Introduction

Forefficien tdiseas e control,i twa sconsidere d importantt oobtai n information aboutth eseasona lvariation s insporulatin g capacityo fan dinfectio nprobabilit y by C. coffeanum, inhabiting coffeebranches ,i nrelatio nt orainfal lan dtimin go f fungicideapplication .Fo rthi spurpos etw oarea si nKeny awer eselected ,namel yi n theEmb udistric to nth esouth-easter nslope san di nth eMer udistric to nth e north-easternslope so fMoun tKeny a (Fig.3) .I neac hdistric t foursite so f Africansmallholder swer e selectedwit h different levelso fCB Dan dsprayin g his­ tories.Th edetail so fal lsite sar eliste di nTabl e 10.I nadditio nt oth e in­ formationi nthi stable ,th efollowin g comments shouldb emade :

-I nth eEmb uare ath efou rsite swer eal lsituate di nRunyenjes ,a villag eabou t 25k mfro mEmbu-town .Th erainfal li nth eobservatio nare awa srecorde di nth e village.I nth eMer uare ath efou rsite swer eal lnea rth esubstatio no fth e CoffeeResearc hFoundation ,wher eth erainfal lwa srecorded . -Th eEmb udistric tha sclimati ccondition salmos tsimila rt otha to fth ecoffe e growingarea saroun dKiambu .Th eclimati c conditionso fMer udiffe r fromth e resto fth eEas to fth eRif ti nthi srespec t thatth emai nrain yseaso nusuall y occursi nSeptembe rt oNovembe r (HUXLEYe tal. , 1969). -Th espra yprogramm eapplie do nsom eo fth eselecte dsite s (Table10 )coul db e modifiedt oa certai nexten ti n196 8an d196 9i norde rt oinclud eno tonl yth e pre-rainsprays ,bu tals ospray sapplie ddurin gth erains .Thes emodification s madei tpossibl et ocarr you tobservation so nth efavourabl e effectso fth e protective actiono ffungicides ,a salread y indicatedb yth eresult so fth e spray timing trialsa tKital ean dMer u (VERMEULEN& PATWA ,1966 ;reprinte do n p. 66-71)an dals oeffectivel ydemonstrate db yMULLE R (1964)i nCameroon .

Thedat acollecte di nth eperio dMarc h 1967t oJul y196 9covere drainfall , sporulating capacity (4.7),pathogenicit yo ngree nberrie so fconidia l suspensions obtained frombranche s (4.7)an dthree-monthl yvisua l scoringso fCB Dincidenc e inth efiel do nlabelle dbranche s (4.8). Iti sclea rtha ta selectio no fcoffe e sitesi nregion smor e thoroughly researched,woul dhav eyielde dmor e standardized informationo nCBD .

64 Table 10:Detail s of selected smallholding sites in theEmb uan dMer u areas ofKenya .

Site . Altitude Coffee Spraying , TLocation . .. Spray history and schedule code inm cultivars code

Embu 1500 French -befor e 1969:2- 3 copper spraysbe - Cu-3 Mission, fore therain s S.L. 34, - 1969:2 coppe r spraysbefor e therain s Kent 2-A copper sprays during and after therain s

Meru 2000 S.L. 34 -befor e 1968:2- 3 copper sprays before Cu-3 the rains - 1968/1969:2 coppe r sprays before the rains - 1969:captafo l sprays during therain s

Embu 1650 French before 1968:2- 3 copper sprays before Cu-2 Mission, the rains S.L.34 , 1968/1969:n o application of fungicides Kent

Meru 2000 Kent -befor e 1968:occasiona l copper sprays Cu-3 before the rains - 1968/1969:coppe r sprays beforean d during therain s - 1969:captafo l spraysdurin g the rains

Embu 1600 French before 1964:pre-rai n copper sprays Cu-1 Mission, after 1964:n o fungicides applied S.L. 34

Meru 2000 S.L. 34 -befor e 1964:n o fungicides applied Cu-3 - 1967:2 coppe r sprays before therain s - 1968:2 coppe r spraysbefor e and during theFebr .t o June rains - 1969:2 coppe r sprays and 2captafo l sprays during therain s

Embu 1600 French no application of fungicides ever Cu-0 Mission, S.L.34 , Kent

Meru* 2000 S.L.34 , no application of fungicides ever Cu-0 Kent

Legend: CU-3= recommende d programme of 4-6 coppe r sprays per annum. 2 =programm e of copper sprays either just commenced or terminated duringobservations . 1= histor y of copper applications. 0 =n o copper or other fungicide applications ever.

Observations started inJanuar y 1968.

65 6.2A shor t noteo nsom e coffee spray timing trials against coffee berry disease

by H.VERMEULE N Plant Pathologist and N.T.PATW A Scientific Assistant Coffee Research Station,Ruiru ,Keny a (reprinted from Kenya Coffee, 31, 368, 343-345, 1966)

Introduction

Themai nproble mo fth ecoffe e industryo fKeny ai scoffe eberr y disease(CBD ) causedb yth efungu s Colletotrichum ooffeanum Noack.BOC K (1963)foun d thatfun ­ gicides basedo ncoppe rwer emor e efficient thanothe r fungicides incontrollin g CBD.H eals o establisheda spra y timing schedulefo rbot hEas tan dWes t Rift areas,havin g concluded that therewa sfo rbot h regionsa critica lperio di n February- April ,whil efo rWes tRif t areasa secon dperio do fsprayin g during June-Octobe rwa sals oadvisable . AsCB Dha sno wsprea d intovirtuall yal lcoffee-growin g areaso fKeny aan d the losses causedb yth edisease s have increased steadily duringth elas t fourt o fiveyears ,i tha sbee nnecessar yt oobtai nmor e detailed informationo nspra y timingfo rthos earea swher epreviousl yCB Dwa so flittl e importance. Thispape rsummarise sth eresult so ftw ocoppe r spraytimin g trials carried outi nth eKital ean dMer uDistricts .

Spray_TimingTria l -_Kitale

This trialwa sstarte da tth ebeginnin go f1965 .I tha dt ob ediscontinue da tth e endo ftha tyea rbecaus eo fth eprevalenc eo fsuberise dberrie swhic h interfered withth eassessmen to fth eincidenc eo fCBD . The following treatmentwer e applied ina randomise d blockdesig nwit h eight replications: T-| September-October ,tw osprays ,a ttwo-weekl y intervals. T2 February- March , four spraysa ttwo-weekl y intervals. T-]+ T 2Tw ospray si nSeptembe r- Octobe ran dfou r sprays inFebruar y-March . C Control-unsprayed. A cuprous oxide fungicide ('Perenox')wa sapplie da ta rat epe racr eo f 10l bfo rth efirst ,7 l bfo rth esecon dan dthir dan d1 0l bfo rth efourt h spray inth eT 2treatmen tan da t1 0l bpe racr efo rth eT- |treatment .Al lspray swer e appliedwit ha Coope r PeglerNo .9 bucke tpum pprovide dwit htw olances .

66 The incidenceo fCB Di nthi s trialwa sassesse db yweighin g separatelyth e healthy andaffecte d ripeberrie san db ycountin g thenumbe ro fhealth y andaf ­ fected greenberrie so nthre e labelledbranche spe rtre e from three treespe rplot . Because theoccurrenc eo fsuberise d berries increased greatly aboutsi xt oeigh t months afterth etria lbegan ,i tbecam emor ean dmor edifficul t toasses s thein ­ cidenceo fCBD .However ,som eresult so finteres twer e obtained. The datafro mth egree nberr ycount s arepresente d inTabl e 1,whil eth e yieldsar egive ni nTabl e 2.Althoug h thedat acove ronl y about fourmonths ,the y suggest thatsprayin g inFebruar y- Marc hwa smor e effective than spraying onlyi n September-October ,an dtha ta combinatio no fthes etw otreatment swa smos tef ­ fective.However ,i tshoul db enote d thateve nth eT- |+T 2 treatment gavea poo r controlo fCB Dsinc e therewa sabou ta 4 0pe rcen t losso fcrop .Whethe ro rno t thisca nb eattribute d inpar t toth eoccurrenc eo fsuberise d berries isstil lun ­ known.

^Er§Z_Ti!!!inSTrials,_Mer u These trialswer e started in196 2i nthi s localitybecaus eMer udistric tha sa climaticpatter ndifferen t fromth eres to fth eEas tRift .I twa sals ohope d that these trialswoul dprovid e informationo nth eeffectivenes s ofcoppe r sprayingo n a smallholding scale.Th etrial swer e concludeda tth ebeginnin g of196 Safte r adequate informationha dbee nobtained . Fourdifferen t experimental siteswer e selectedan do nthes e sitesth etrial s were laidou ta srandomise dblock swit hth efollowin g treatments: T-| September-October ,fou r sprays attwo-weekl y intervals. T2 February-April, fourspray sa ttwo-weekl y intervals. T-|+ T 2 Fourspray s inSeptembe r-Octobe rplu s four sprays inFebruar y- April . C Contro1-unsprayed. Cuprousoxid e ('Perenox')wa sapplie dwit h 'Saval'hydrauli c knapsack sprayers ata rat eo f1 0l bpe r10 0gallon spe racre .Th eyiel d ofeac ho fth e plotso nal lsite swa srecorde db yweighin g allrip ehealth y andaffecte dberries . Furthermore,health yan daffecte d greenberr y countswer e carriedou ta tfort ­ nightly intervalso nbranche so nuntreate dplot s adjacent toth efou rtrials . The informationobtaine d aftertw oan da hal fyear swa sdifficul t tointer ­ preta sto oman y different factorsha dbee n included. InTabl e 3th eresult sar e presented.Th eanalysi so fth eyiel d resultsa tfou r siteswa scarrie dou to nth e combined twoseason s inbot h 1963an d1964 .Th esecon d cropo f196 2give s ineffec t the totalyiel do ftha tyea ras ,du et oth eflood s intha tyear ,ther eha dbee nn o first-seasoncrop .

67 Inth ethre eanalyse s carriedou to nth ethre eyear syiel dresult s fromfou r differentsite sth etreatment sbetwee nsite s showedn odifferences .I nal lth e threeyear sa tal lfou rsite sth eT- |+ T 2treatmen t gaveth ehighes tyiel do fclea n coffee followedb yth eT- |treatmen t (Table 3).Sprayin g only inFebruar y-Apri l (T2)gav ea lesseffectiv econtro lo fCB D (Table 3).Thes eresult s showtha tth e maintim efo rsprayin gagains tCB D inth eMer udistric t isth eSeptembe r-Octobe r period,wit ha ver ybeneficia l effecto fth eadditiona l February-Apri lsprays . Thespra y timingschedul e inth eMer udistric t thereforediffer sb ysi xmonth s fromth eEas tRif tregions ,wher e thebes tcontro lagains tCB D isachieve d in February-Apri lwit ha nadditiona leffec to fSeptembe r-Octobe rspray s (BOCK, 1963).

Summary

1.I nth eKital edistric tth emai nsprayin gperio d shouldb eFebruar y-April ,bu t additionalspray s inJun e-Octobe rar eadvisable .

2. Inth eMer udistric t themai nsprayin gperio d shouldb eSeptembe r-October , witha nadditiona lCBD-controllin geffec to fFebruar y-Apri lsprays .

Acknowledgments

TheKital etria lwa s carriedou tb yMr .R.A .Reynolds ,Mr .J.K .Ammond san dth e field staffo fth eKital eCoffe eResearc h Sub-station.Mr .J.B .Jacob ,Kitale , kindlyoffere d theopportunit y tocarr you tthi s trialo nhi sestate . TheMer utrial swer e startedb yMr .B.F .Hange ran dMr .I.D .Firma nan d carried outb yMr .J.H.G .Waithaka ,Mr .J.W.E .Njagi ,Mr .S.E .Ngati aan dth e field staffo fth eMer uCoffe eResearc h Sub-station.Fou rprivat e farmerswer e kindenoug h toallo wu s tocarr you tth etrial so nthei rplots .

Reference

BOCK,K.B . (1963). 'Thecontro lo fCoffe eBerr yDiseas e inKenya' . Emp. J. exp. Agria. 31,122 ,97-107 .

68 Table1 (reprinted paper Kenya Coffee, 31, 368, 343-345, 1966)

KITALE CBD SPRAY TIMING TRIAL

Results of thegree nberr y counts (figures in the table aremean s of 72branches )

Overall percentage number diseased berries

Control Counts Dates in 1965 T,+T Tl T2 2 (unsprayed)

1 5 June 0.88 0.20 0.10 0.30 2 21 June 5.77 1.87 1.11 2.25 3 5 July 5.81 1.57 1.18 1.39 4 19 July 5.94 2.51 1.65 2.56 5 4 August 7.71 3.21 1.61 5.11 6 20 August 9.05 3.35 1.28 7.41 7 3 September 10.37 4.63 1.78 8.87 8 17 September 9.33 4.32 1.29 7.99 9 30 September 9.29 4.11 1.54 6.96

Total 62.15 25.79 11.5 2 42.84 Mean 6.91 2.86 1.28 4.76

T]= Sept.-Oct. , 2spray s T2 =Febr .-March ,4 sprays Tl+T 2 = Sept.-Oct. , 2 sprays +Febr . •March,4 sprays C =unspraye d control

69 .-i -a in c^ LO \D CO T-i /—V 4-i a) 60 o -H .M FH ^^ 0) 60 T3 a) co a) 60 CO C co IJ CU CU a> a en + 1 CD T3 P.

CD 60 •H ^ >H ^

CU > 60 T3 H o 3) CU >,^N O V — 53 U B t< C C co U CO CNO a> cu cu cu H o CO .C S II 60 H 'H o ^ + CU H

CU 60 •H Ji >H -_- CO M CU CN'-s P. co 60 T3 H en CO 3 cj 1) >,^ -a •u en u c v ° CN B co >j co CNO cu cu cu v£> v£> - 1 IS o o 4-1 CO ^1 C 1) M VJ C CO ij CO CJ CU CU QJ CU + .S> S m en co m M o co ,e S <<*-K, H -a O cu CU T3 C/> •H CU 60 H ~_- p- Ss c_> 0 o •» cu w 60 T3 n. 4-> t? hJ « « >,^s a, -C U CO U C H *~'

o CU 60 TJ- cO CU >1 /—s 4J CO SJ J3 C co u CeO co 60 CU CU cu 0) H CJ CO •C + 1 a a U -i-l CJ r, CU T3 o CM •C T3 i—I -~v OJ 60 •H J*S J2 >H w CO H CU 60 -a CO cu >1 s-N 4J CO u a co u CcO CO CJ\ vO V o CU CU CU CI) CJ CJ CO .a R + I H O O U -H c) CU T3 PH

hn CI CM O C •H u"> 4-i en CO I-I Cti 4J CU ^!c^) OJ C C .c*) •U — o 70 H CO H S PH T3 Table 3 MERU CBD SPRAYTIMIN G TRIALS (reprinted paper Kenya Coffee, 31, 368, (a)J.962_viel d 343-345, 1966) Meanyiel d over four siteso f each treatment Average Cleancoffe e in Treatments yield in cwt./acre kg/plot Tl • 90.80 7.99 T2 76.30 6.74 T, +T2 93.60 9.49 Control 66.90 5.90 A two-way table showing the least significant difference (kg cherry per plot) Control Tj difference 66.90 90.80 + 23.90 S.E.= + 4.87 T2 T]+T 2 L.S.D. =+ 9.69 p= 0.0 5 76.30 93.60 + 17.30 + 12.84p = 0.0 1 difference + 9.40

(b)I963_y_iel d (the two cropping seasons combined) Meanyiel d over four sites of each treatment Average yield in Clean coffee in Treatments kg/plot cwt. /acre Tl 54.35 4 80 T2 52.53 4 64 T, +T2 61.20 5 40 Control 39.64 2 61 A two-way table showing the least significant difference (kgcherr y per plot)

Control Tl difference 39.64 54.35 + 14.71 S.E. = + 4.27 T2 Tl +T2 • S.D. = + 8.50 p = 0.5 52.53 61.20 + 8.67 + 11.26 p = 0.1 difference+ 12.8 9 +6.85

(c)J.964_y_iel d (the two cropping seasons combined) Meanyiel d over four siteso f each treatment Average yieldi n Clean coffee in Treatments kg/plot cwt./acre Tl 53.42 4.71 T2 51.69 4.55 T, +T2 74.38 6.56 Control 43.84 3.86 A two-way table showing the least significant difference (kg cherry per plot)

Control Tl difference 43.84 53.42 + 9.58 S.E. = +_3.2 6 T2 T] +T2 L.S.D. =+ 6.49 p= 0.0 5 51.69 74.38 + 22.69 + 8.60 p= 0.0 1 difference +7 . +20.9 6

N.B. The average yield inkg/plo t isconverte d toclea n coffeeweigh t expressed incwt./acre .

71 6.3Sporulatin g capacity inrelatio n torainfal l andfungicid e applications

The dataobtaine do fth etota lS.C .o fth ebranche so fth esite sa tEmb uan dMer u aredepicte d inFig s9 an d10 ,respectively , togetherwit hth erainfal l figureso f eachare aan dth ecoppe r spray scheduleapplied .Th eseasona l fluctuationso fth e totalS.C .ar esimila rt othos epreviousl y reported forth eKiamb uare ab yNUTMA N andROBERT S (1961, 1969a),i.e .th eS.C .produce dpeak s inconidia l production beforeth eonse to fan ddurin gth erains .Th evariou sS.C .level so fth eindivi ­ dualsite s showed,however ,marke ddifference s causedb yth edifferen t spray regimes.Comparin gth eS.C .dat a fromEmb uan dMer ui tshoul db enote d thatth e Embu figures showa gradua l increase towardsth een do fth eobservatio nperiod , viz. thebeginnin go f1969 ,whil ea tMer uth elevel so fS.C . remain loweran dmor e regular.I nbot h locationsth ehighes tS.C . figuresar ethos efo rth eunspraye d controltreatments .

n JI I JJ J SPRAy SCHEDULE J J J nun T f T t

300- EMBU Sporulating capacity, internodes 7-\U

§200-

o /' A d t &j$ Mm SO 2(0 V) #200 ID

§ 100 E E MAMJ JASONDJ FMAMJ JASONDJF 1967 *• 1968 —— -oSit e A --SiteC -SiteB •••»SiteD

Fig. 9: Sporulating capacity data (sensu NUTMAN& ROBERTS) , rainfall figuresan d spraying schedulefo rcoppe r fungicides (indicatedb yvertica l arrows)i n the Embu observation area. 72 TheS.C .dat ao fth eindividua l internodes,viz .internode s7-1 4o feac h sitear eno tgive n separately,a sth epresentatio no fal lthes edat awoul d have beento odetailed .The yal lsho wth esam e seasonal rhythmwit hth ehighes tnumber s ofconidi aproduce dpe rairV h forth einternode s 10, 11an d12 ,a sreporte db y MJTMAN andROBERT S (1969a)fo rmateria l obtainedi nth eKiamb uarea .

Coptafol sprays in n n n n II mi ii • t t • • •: a o op!1?

-1967- 1968- -1969- o^_o Sjte A D.—a siteC -—• Site B » •»Sit e D

Fig. 10:Speculatin g capacity data (sensu NUTMAN& ROBERTS) , rainfallan dsprayin g schedulefo rcoppe ran dcaptafo l fungicides (indicatedb yvertica l arrows) inth eMer u observationdata .

6.4Pathogenicit y ofconidi a producedo nbranche s inrelatio nt osporulatin g capacity

Thepathogenicit yo fconidi awashe dof fbranche swa sassesse dfo reac ho fth eeigh t internodes separately.N oinfection so ngree nberrie so fsusceptibl e coffeevari ­ etiescoul db eobtaine dwit hth econidi asuspension so finternode s9-14 .Fro mth e sprayed sitesth econidia l suspensionso finternode s7 an d8 onl ygav emarke din ­ fectionso ngree nberrie susuall y duringo rshortl y afterth erains .I nFig s1 1

73 and 12thes e findingsar erepresente dfo rEmb uan dMe m respectively,combinin g thecroppin g cycleso nth etree san dth eS.C .dat ao fth eintemode s 7an d8 wit h theirpathogenicit yo nberries . Thepathogenicit yo fconidia l suspensions obtained from internodes 7an d8 shows fluctuations relatedt orainfal lan dfungicid e applications.Th eeffec to f the lattero nth epathogenicit y isver ymarke do nth esite s subjectedt ocoppe r sprays,viz .Embu ,sit eA an dMeru ,site sA an dB (seeals oTabl e 10).Th e im­ plicationso fthes e findingswil lb ediscusse di ndetai li n6.6 .

"g ENS.HG S,R B FL.PH. EX S ,EN S .HGSRB FL.PH.EX S .EN S. | FL.PH.EX S. EN S .HGS.RB FL.PH,EX S .EN S .HGS.RB FL.P H,EX S •s too V j» Pathogenicity of c ;! bark conidia

1) / i internodes 7+8 Q. I 50. O 1 c *~\ ! \ YJv-*

I I 1 1 1 I T—1 1 1 1 r »&£$

585 (95 ..•• / ••...360.. =2= •••" 300 —° Site A •- SiteC J* —• SiteB ••» Site D 1 »•• f' I EMBU Sporulating capacity Si—t 3 O internodes 7+8

•1967- -1968- -1969-

Fig. 11:Sporulatin g capacity data of internodes 7+8, pathogenicity of the conidia of internodes 7+8 and the sequence of cropping cycles inth e Embu coffee area (see text 2.1.3.3, 4.7,4.8) .

74 ENS.HGS.RB FL.PH . EXS . ENS .HGS.RB FL.PH , EXS . ENS .HGS FL.PH. EXS ENS .HGS.RB FL.PH. EXS . ENS .HGS.RB FL ,PH .EXS

Pathogenicityo f bark conidia. 50- p. internod.es 7+8 Q. s ,pN I \pKJV-"" 1'\L-9 V/*/ V

200-

MERU Sporutating capacity. internodes 7+8 E 100- u g -a 'c ^jS^j^^;,^^^ ^ MAMJJASONDJFMAMJ JASONDJFMAMJ J 1967 •« 1968 •« 1969 -° SiteA -•<• SiteC -• SiteB ~» SiteD

Fig.12 : Sporulating capacity data of internodes 7+ 8,pathogenicit y ofth e conidia of internodes 7+8 andth esequenc e of cropping cycles inth e Meru coffee area (seetex t 2.1.3.3, 4.7, 4.8).

6.5CB Dincidenc e inth efiel d onlabelle d branches

The datao nth ethree-monthly ,visua lCB Dscoring so nmarke dbranche sprovide d use­ ful informationo nth eincidenc eo fth ediseas e inth esmallholdin g sites.I n Table 11 these resultsar epresented . Themai ncrop ,picke d inOctobe rt oNovember ,i nEmb ui striggere db y theMarc ht oMa yrain san di sfollowe db ya smalle r crop,picke di nFebruar yt o March.I nMer uth erevers e occurred.I ngenera lth epotentia l crop,viz .th ecro p potentiallypresen to nth etree s shortly afterth eterminatio no fth eflowerin g and subsequent fruit set,i n196 7o nsite s subjectedt oa sprayin g programme (Embu A,B an dMer uA ,B o rC )o rwit ha recen thistor yo fspra y application (EmbuC ) washighe r thano nth eunspraye d site,viz .Emb uD .Th ehig h incidenceo fCB D duringpart so fth ecroppin g cycle resultedi nconsiderabl e lossesan dultimatel y

75 * o o *I oI *I oI O I I I I Q I * o 0) t—\ 3 I I I I M o CJ o * o * J3 u I I I I CO Hi —. o * o * H H .- CN * *I oI co O CO 00 4-1 o u c I T oo •H r*- o O o o l-l ^o I I I u o> O a O() 3 I I I CO «^ CJ * O o J3 I I I CO CJ * * o u c«a U p. i) ccl ?! o o rH I I 7 f I o co T) o o 3 O f T 7 I 7 7 T CO •H CJ I o CJ •H U 3 I I cu CJ o o > a ? T I I C * o CO CU o * t—>,t J 2 X! 4J * o * o 4J I I I i I C C * * O •H I I I i a S o o cu 1 CU 3 I I o I I co (U 4-> CJ o o o CD •i-i I I I u U CO o J= 4-> u a CU o # A # M M-l a I I I I A I CJ O o * * A O 6-8 oo CO I I I A I O c CO CU <: i o CN O A CN 4J x: 3 I I I I A I s r-l o CJ o — A O CJ 3 a I I o I A I o a CO CO * OJ A CN a

r^ r-~ oo eft c^ eft cn 0\ ON I I I I •a CO cn ^D co \o o\ CN CO \£> o o o o — o o J3 a I I I I I I at CO o> o — r- 00 H o — — CN CU •J

76 thecro pharveste d fromth evariou s sprayed sites didno t differ from theyiel d onth eunspraye dplot . The samecoul db e observed in 1968 and 1969:highe rpotentia l crops on sprayed sites,bu t serious lossesb y CBD reduced theyield . Inthos e caseswher e farmers,viz .i n196 8Mer uA ,B andC ,Emb uB and in 1969als oEmb uA , switched tospra y applications during the rains,a marke d decrease ofCB D incidence (Table 11)coul db eobserve d especially inth e susceptible,soft-gree nberr y stage.Th e applicationo ftw oOrth oDifolata n sprays at theen do fth eobservatio nperio d in Meru furthermore improved theCB D controlling effects ofth ecoppe r sprays applied during the rains.

6.6 Discussion

The results ofth e spray timing trials inKital e andMer uprovide d at that time (1966)alread y indications that copper fungicides acted asprotectant s onberrie s (MILLER, 1964). Themos teffectiv e regime inth e two areasprove d tob e thecom ­ bination ofpr e andpost-rai n sprays.Considerin g thepersistenc e andredistri ­ butiono fcoppe r fungicides (Chapter 7)an d thedifferen t levels of susceptibility of thegrowt h stages ofcro p to C. aoffeanwn, the following explanation seems valid: thepre-rai n copper spraysprovide dprotectio n to the susceptible flowering stagean dexerte dmos t likely still somediseas e controlling effects onth eex ­ panding berries six:t oeigh tweek s after application;th epost-rai n sprays pro­ tected the lastpar to fth esoft-gree nberr y stage and thebeginnin g ofth ehard - greenan d ripeningberr y stages (Table 2;Fig .6) .Eac h treatment separately,viz . eitherpr e orpost-rai n sprays,coul dno tprotec tmor e thana par to fth e crop cycle,respectivel y the flowers and thebeginnin g ofth e expanding greenberrie s or theendosper m stagean d theonse t of theberr y ripening (Fig. 6).Althoug h the results of these timing trials infac t already implied theprotectan t action of thefungicide s (Chapter 7),th e general recommendations remained inthos eyear s aimed at theeradicatio n ofbar k inoculum of C. coffeanum. The climatic conditions inEmb u andMer uonl y differ to theexten t that the majorflowerin g occurs respectively inEmb u inMarc h toApri l andi nMer u in October toNovember . Inbot h areasusuall y asecon d flowering occurs respectively inOctobe r toNovembe r andMarc h toApri l and thepresenc e ofprecedin g crops,a s shown inFig s 11 and 12,seriousl y affected thenewl y initiated crop,becaus e the smallholderswer eno twillin g to sacrifice thesmalle r cropb y stripping itof ft o prevent re-infection ofth emai n crop (MULLER,1964) .Th e overlap ofth etw ocrop s therefore causedconsiderabl e losses,especiall y inth e susceptible stages,viz .

77 flowers,soft-gree nberrie s and toa lesserexten t therip eberrie s (MOGK, 1970). Although affected ripecherrie scoul d stillb epicked ,the ywere ,while o nth e trees, importantdiseas e transmitters (GIBBS,1969) . GIBBS (1971) foundtha ti n estate coffee therol eo foverlappin g crops intransmittin gCB D isfrequentl y less important,whe nadequat e sprayapplication sprotec tbot hth eol dan dth ene wcrop . Under thesmallholdin g conditions,however ,spray s areofte napplie d lessfre ­ quently andefficientl yan dunde rthes e conditions cropoverla pofte nresult si n seriouscro p losses. Inth eEmb uare ath eS.C . figureso fth esite swit h littleo rn ocoppe rspray ­ ingshowe d agradua l increasedurin g theobservatio nperio da scompare dt oth e S.C. ofth ecoppe rspraye dplo t (A).I nMer uth eS.C . figureswer egenerall y lower thani nth eEmb uarea ,du epossibl y alsot oth ealtitud eeffec t (NUBIAN& ROBERTS , 1969a),whil eth e continued applicationo f fungicidal sprays (A,B ,C )kep tth e totalleve lo fsporulatio nsuppressed . TheS.C . findingso f individual internodeso feac h siteshowe d thesam e seasonal rhythma snote db yNUTMA N andROBERT S (1969a).Th e internodes 7an d8 gavemuc h lowerconidia lcount s thanth e internodes9 , 10, 11 and 12,while inter ­ nodes 13an d 14gav e also lowerfigure stha nth e internodes 9-12.A sn o infec­ tioncoul db e inducedwit h conidial suspensionsderive d fromth e internodes 9,10 , 11, 12,1 3an d 14,i ti scertain ,als oconsiderin g theanatomica levidenc e (VERMEULEN, 1970a), that C. aoffeanvm occupiesonl ybranc hsection swit hearl y signso fbar kmaturatio nan d ishardl y orno tpresen t inth eolde r internodes,viz . the internodes9-1 4 (FURTADO,1970 ;VERMEULEN , 1970a).Th e internodes,containin g C. coffeanum, are stillrelativel y smoothan dwithou tan y surfacecrack san dcrev ­ ices (2.3,4.7 ) and therefore it islikel ytha tpesticide scanno t remainattache d tothi ssurfac efo ra longperio do ftime ,especiall ydurin g therains .Hence , thissuggest s that fungicideswoul db ewashe dof fmor e rapidly fromthes esmoot h surfacestha nfro mth ecracke dan droug hsurface so folde r internodes,permittin g thecolonizin g C. eoffeanum inth eyoun g internodes toproduc eo nth ebranc h surfaceconidi aabl e toinfec tcro pmor e readily thato nothe rinternodes . Thepathogenicit y ofth econidia lsuspension sobtaine d fromth e internodes7 and8 o ndetache dberries ,a sdepicte d forEmb u inFig .1 1an dfo rMer u inFig . 12, showed forcertai nspraye do rrecentl y sprayed sites definitepeak s inpatho ­ genicitygenerall yjus tbefor eo rafte r increaseso fth eS.C . onth ebranc hinter ­ nodes 7an d 8.Th epathogenicit y ofth e conidiaobtaine d fromsite sB andC a t Embugraduall y decreased inconjunctio nwit h thedecrease dnumbe ro fspra yap ­ plications,whil e theS.C . figureso fthes e sites increased inth eobservatio n •period.O nth eMer usite s thepathogenicit y ofconidi a fromsit eA (copper sprays),

78 B (coppersprays )an dC (copperapplication s started in 1967)wa s inmos t cases highertha nth epathogenicit y ofconidi a fromth eunspraye dcontro l in196 8an d 1969,especiall y forth etw o first sites (Fig. 12). Itshould ,however ,b ekep t in mind thatth e recordedpathogenicit ywa sbase do nlaborator y findingsunde r ideal conditions forth epathogen . Inth e fielda mor e continuous infectivity ata lower level -du e toweathe r conditions (WOODHEAD,1968 )- seem slikel y tooccu rbu ti t may,however ,b e assumed that inth e fieldth e increased releaseb y thebar ko f C. eoffeanum conidiaals o takesplac e during andshortl y after therains . InEmb uth eeffect so fth e inoculumo f C. eoffeanum frombar kwer emos tmarke d in196 7an d thefirs tthre emonth s of 1968fo rsite sA ,B an dC (Fig. 11). Itseem s certaintha tsusceptibl e croppresen t atth epeak s ofpathoge nproductio n in thebark ,wa s infected attha t time,especiall ywhe nrainfal lwa s lightan dpro ­ vided theopportunit y forbar kwashing s tob edeposite d forperiod s sufficiently long forinfectio no nberrie s (WALLER,1972a ;WOODHEAD , 1968). In196 8onl yon e more,ver ymarke dpotentia l infectionpea k inth e C. eoffeanum bark production occurred onsit eA (continuous copper spraying).Thi s sourceo finoculu mmos t likely affected attha ttim eth ene w flusho fflowers .I nth eMer u areath e pathogenicity of C. eoffeanum inth ebar kcoul db e established during orjus t after therain so nthos e sitesspraye dwit h fungicides.Agai n itma yb e assumed thatcro p infection inth efiel doccurre dunde r conditions favourable fordisseminatio no f C. eoffeanum bark conidia toneighbourin g crop. The effecto fapplication s of fungicidesdurin g therain s in196 8 (MeruA , C)an d in196 9 (EmbuA ;Mer uA ,B an dC )coul db egauge d fromth eCB Dscorin g data inTabl e 11wher e agradua l decrease innumber so faffecte dgree nberrie swa sob ­ served.Th eactua l actiono fberr yprotectio nb y thefungicide ,a sshow nb yMULLE R (1964,1968 )i nCameroo nan d laterb yGRIFFITH Se t al. (1971)i nKeny awa s thus also found inth eEmb uan dMer uareas .

79 7. FUNGICIDE SCREENING TRIALS

7.1SCREENIN G FUNGICIDES FOR CONTROL OF COFFEE BERRY DISEASE IN KENYA SUMMARY Fourteen out of4 3fungicide s wereselected , after laboratory screening, for field trials against coffee berry disease. Each fungicide was compared with a commercial cuprous oxide formulation and an unsprayed control,b yassessin gconidia lproductio n ofbearin g coffee branchesan d by recording the percentage of healthy ripe berries picked. In assessments of fungicidal efficiency, healthy ripe berry counts gave more consistent results than those obtained by measuring conidial production. The results from berry counts indicate that in these experiments Ortho Difolatan (BSI 'CaptafoP) was equal, but not significantly superior, to the standard formulation used in estate spraying. The annual rhythm of conidial production seems to have changed over the past five years. The significance of this is discussed in relation to the control of coffee berry disease.

Prior to 1961, the coffee berry disease (CBD) ofCoffea arabica L., caused by a form of Colletotrichumcoffeanutn Noac k {Glomerellacingulata (Stonem.) Spauld. and Schrenk), was responsible for serious crop losses in coffee areas situated above 5,500 ft (Nutman & Roberts, 1961), but since then, following abnormally heavy rains, CBD has rapidly spread into lower districts, where it now causes heavy losso f crop. Bock (1963) reported that in the higher altitude areas copper fungicides gave a higher degree of control of CBD than twenty-five non-copper formulations. Al­ though an organo-mercury fungicide was even more effective than copper formu­ lations, this compound produced phytotoxic and harmful residual effects (Bock et al., 1958). The present recommendations of four to six sprays per year, using either a two per cent Burgundy mixture or 10 lb per acre of a 50 per cent copper formulation, do not seem to give adequate control. CBD caused increased losses in 1964. Although these were certainly attribu­ table in part to wrongly-timed copper sprays (Bock, 1963) and to inefficient application, it was felt necessary to resume research on fungicides in an attempt to find a compound superior to the copper formulations. It was also desirable to find a fungicide with an eradicant action, which would destroy the mycelium of C. coffeanum inhabiting the bark tissue of bearing branches of the coffee tree (Nutman & Roberts, 1961).

MATERIAL AND METHODS Laboratory screening Five chemicals and 43 fungicides (Table 1) were tested by comparing their ability to prevent the germination of conidia of C. coffeanum, using the 'slide' method described by Nutman & Roberts (1962). It was fully realized that this method, * Present address: E.A.A.F.R.O.-CBD-unit, Nairobi, Kenya. 80 like any other laboratory screening technique, would not necessarily give results closely comparable to any future field performance. It was, however, at least considered likely that in a relatively short period the unpromising compounds could be discarded, because of their inability to inhibit germination ofC. coffeanum conidia invitro.

Table i. Trade names orcode-numbers of tested compounds Cadmium chloride crystals Dodine acetate Cadmium sulphide Tricarbamix Cadmium sulphate pure Ortho Phaltan (= Folpet) Mercury bichloride chemicals Ortho Difolatan (BSI 'Captafol') Stannous citrate DAC-2787 Perenox Olin-1763 Shell-75 Olin-1562 Fungex Halophen (C-4) Coprantol Venturicidin Colloidox Allisan BCO-75 Verdasan Cupravit-Blue Manzate-D DuTer Zerlate (= Ziram) Brestan Morestan Dyrene Urbacid Glyodin Tuzet Acti-dione Antracol Blasticidin Euparen Fungicide 328 Fungilon ^Bayer AG* Captan Pomasol EL-211 Lonacol Lead arsenate Bayer No. 5511 Sabithane-M Bayer No. 5467 Dithane M45 Bayer No. 5468

The conidia used in these tests were harvested from lesions on ripe coffee berries picked from SL-34 trees on the Coffee Research Station, Ruiru. The conidia were washed from the lesions after 24 hours incubation at 100 per cent R.H. at room temperature (2i°-22°C). For each test, new ripe berries with CBD lesions were used to obtain fresh conidia ofC. coffeanum. Thin films of potato dex­ trose agar, mixed with known concentrations of fungicides, were prepared on microscope slides and seeded with a io5 conidia/ml. suspension, applied with an 'Aerograph' airbrush. After seeding, the slides were incubated at room tempera­ ture in plastic boxes lined with wet cellulose wadding for eight hours. The conidia were then killed with lacto-phenol-cottonblue and the percentage germination assessed by counting 200-400 conidia. The fungicides were first tested in con­ centrations ranging from 6 to 100p.p.m. , but where afungicid e did not give marked inhibition of germination at 100 p.p.m., a further concentration range of 100 to 1000p.p.m . was tested.

Fieldtrials Two suitable sites were selected in the high altitude area (Kiambu: Njuno and Kiamara Estates) in January 1965, because a spray-timing schedule for copper * Tested by Bayer AG in separate field trials. 81 had already been established for this altitude (Bock, 1963; Nutman & Roberts, 1965). On both sites the trees were 40 years old, multiple stem and of the French Mission variety. Further details are given in Table 2.

Table 2. Details of two sitesselected for trials Shade Estate Altitude trees Interplanting Main flowering Picking Njuno 6000 ft yes no March 1965 June 1965- January 1966 Kiamara 5500 ft no Blue Mountain October 1964 May 1965- & SL-34 & March 1965 December 1965

The incidence of CBD on both Njuno and Kiamara Estates was high and uni­ formly spread throughout the trial areas. Seven different fungicides were com­ pared with a standard copper fungicide and with the unsprayed control on both estates (Table 3). The field trials were designed as randomized blocks, with nine treatments, eight replications (ten trees per plot) and a single row of guard trees between the plots. Fungicides were first applied with a Drake and Fletcher 'Kingston' headland pump and later with a Holden-BSA spraying unit, using handlances at high- volume (200 gal./acre) and high pressure (200-300 p.s.i.) (Bock, 1963). The normal pest control programmes were carried out by the farm management.

Assessmentof treatment effects To assess the effect of the fungicides on the incidence of CBD in the field the follow­ ing methods were used: 1. Each month a random sample of five fruit-bearing branches was taken from every plot. Internodes with the first signs of bark browning, and seven successive internodes, were treated according the technique described by Nutman and Roberts (1961) and by Bock (1963). The number of conidia produced per cm.2 surface area/hr under favourable conditions (approx. 22°C and 100 per cent R.H.) was subsequently assessed. 2. All ripe berries were picked once every every eight to ten days, from four marked trees per plot, throughout the picking season. Before the picking season started all black, infected and old berries were removed from the branches. The total number of picked fruit, including infected and healthy berries, was recorded. The differences between the treatment means were tested for significance, using the Duncan Multiple Range test at a probability level of P = o-oi.

RESULTS Laboratoryscreening Fourteen fungicides were selected for field trials, based on their consistent ability to inhibit germination ofC. coffeanum conidia invitro at concentrations of 100 p.p.m. 82 or lower. Table 3 illustrates the results of the screening of the fourteen fungicides as compared to a standard fungicide.

Table 3. Inhibition of germination at various concentration 1000 500 250 100 50 25 12-5 6-25 Mercury bichloride* ______DuTer -- + + + + + + Brestan — — — — + + + + Fungicide 328 — — — + + + + + Captan ______Sabithane-M --- + + + + + Dithane M45 — — — — — ± + + Dodine acetate — — — — ± + + + Tricarbamix — — d_ ± + Ortho Difolatan _____ DAC-2787 _____ Olin-1763 ± ± ± ± Olin-1562 _____j_ Halophen — — + + + Venturicidin — — — — —

— = no germination __ = less than 50 per cent germination + = more than 50 per cent germination * used as a laboratory standard

Fieldtrials The data obtained from both trials, on the production of conidia per unit of twig surface per hour, indicate that twigs treated with Perenox (cuprous oxide) and Ortho Difolatan tended to have a lower conidia production than the other treatments. Although this trend was at first not statistically significant, it became more apparent in the latter part of 1965. Furthermore, it was found that on the individual sampling dates, a high degree of variation occurred in the number of conidia produced per cm.2/hr between replicates of the same treatment, while a high variation was also found from month to month in the same replicate. The seasonal variation of the mean of eight replicates of the number of conidia produced cm.2/hr. in the most interesting treatments are shown in Fig. 1. The ripe berry counts on Njuno and Kiamara Estates showed that Perenox in both trials and Ortho Difolatan in the Kiamara trial gave the highest per­ centage of healthy ripe berries. The effect of both fungicides in reducing disease incidence became apparent in the Kiamara trial in May-June and remained consistent throughout the year. On Njuno Estate Perenox gave the highest percentage of healthy ripe berries on all picking dates. The total yield data of ripe berries picked on both estates are presented in Fig. 2. On Njuno Estate, DAC-2787 gave the highest number of ripe berries picked corresponding to a yield of 10-3 cwt/acre, followed by DuTer (9-3cwt/acre) , but their respective proportions of healthy berry, 50 and 38 per cent, were lower than with Perenox (6-5 cwt/acre total yield and 74 per cent healthy berries). On Kiamara Estate Ortho Difolatan gave both the highest number of ripe berries

83 3 o

3 •a o

'a

£ 3 G

Jan Feb March April May June July August Sept Oct Nov Dec

.3 c 5 = B5.S -1— ' JL • —f „, | — I L. Control Njuno Control Kiamara Perenox Njuno A 4 Perenox Kiamara X X Ortho Difolatan Fig. i.—The seasonal variation of the number of conidia produced on branches per cm2/hr in some treatments in the Njuno and Kiamara trials. picked (13-7 cwt/acre) and the highest percentage healthy ripe berries (75 per cent), followed by Perenox with a total yield of 11-9 cwt/acre and 72 per cent healthy ripe berries.

DISCUSSION The laboratory screening of fungicides using the slide technique has inherent limitations. With this method only the inhibition of the conidia germination is being assessed in vitro, but the far more complex conditions in the field after spraying cannot be evaluated. It was therefore unlikely that the laboratory results, covering only one aspect of fungicidal action, would be closely comparable with field performance. In spiteo f this,i twas ,however , considered tob e the only method available at short notice for selecting more promising fungicides for inclusion in field experiments. The annual cycle of conidial production for 1965 differed very much from the cycle found by Bock (1963) for i960. His data showed a marked increase in

84 Njuno Estate Kiamara Estate

o X •s

c

•?.

o

Fig. ?.—The total yield data of ripe berries picked from marked trees on Njuno and Kiamara estates. Useful yield is expressed by total column length, while the healthy berry yield is depicted by the black column only.

March-^April, whereas for 1965 there was no such increase. The reason why the expected maximum conidial production in March-April did not occur is un­ known. The difference in this annual conidial production pattern may be a reflection of the change in annual climatic patterns since 1961. Bock stated that there was a critical period for control, and that this was the period immediately before the March-April maximum. It has been suggested that the critical period for disease control has now shifted to December-January, and that spraying in these months would result in a greater effectiveness of the applied fungicide (Nutman & Roberts, personal communication, 1967). It can be concluded from Fig. 1 that spraying in June-July did not decrease the conidial production to a great extent, especially in Njuno Estate. This observation seems to confirm Bock's findings that the critical period for CBD-control is before the peak in conidial production, but more detailed work is necessary. With regard to the fungicidal effect on ripe berries, Bock found 'because of the extreme susceptibility of the ripe fruit, the degree of its infection over this

85 comparatively long period affords a useful andaccurat e index of the effects of fungicide treatments'. The incidence of berry infection should be an indirect fungicidal effect on the production of conidia of the bearing branch, but from the data obtained it can besee n that berry counts gave more consistent results than those on conidial production. During thefirs t eight months ofth e trial periods it was difficult to relate laboratory values on conidial production toth e incidence of field infection; only in the last three months could the data be correlated. It is possible that thedifference s in total yield (Fig. 2) indicate that varying amounts offrui t were lost before the picking season hadended , i.e. inth eearl y fruit setting stage, asfoun d byMulle r and Gestin (1967). This possibility should be investigated in further field trials, by counting all berries on some marked branches to assess the losses ofearl y fruit. The results of the two trials indicate that, of the new fungicides tested, only Ortho Difolatan could be compared with copper in efficiency, although it was not significantly better at the applied fungicidal rate and spray-timing. The favourable side-effects of Ortho Difolatan, namely on the control of crinkle- leaf and the inducement ofbette r growth of the coffee tree, should also be taken into account. However, the lack of control of leaf rust (Hemileia vastatrix B. etBr. ) with Ortho Difolatan implies that, even if subsequent trials confirm the value of this fungicide for CBD-control, some copper sprays will still berequire d forth e control oflea f rust.

Acknowledgements. I wish to thank Dr.D . Kirsch (Farbenfabriken Bayer AG), Dr. F. J. Nutman, Dr. F. M. Roberts and Dr. P. A. Huxley for their helpan d advice and Miss N.T . Patwa, Mrs. A.M .Vermeule n andth e field staff ofth e Coffee Research Station, Ruiru, for assistance in thelaborator y andfield . Mr.S . A. Lagerberg (Kiamara Estate) and Mr.B . M. Aagaard (Njuno Estate) kindly provided theopportunit y to carry outth etrial s on their estates. This paper is published with the approval of the Director of Research, Coffee Research Founda­ tion, Kenya.

REFERENCES BOCK, K.R . (1963). Emp.J. Expl Agric. 31,97 . BOCK, K. R., ROBINSON, J. B. D. & CHAMBERLAIN, G. T. (1958). Nature 182, 1607. MULLER, R. A.& GESTIN, A.J. (1967). Cafi,Cacao, The' 11, 157. NUTMAN, F. J. & ROBERTS, F.M . (1961). Trans. Brit, mycol. Soc. 44, 511. NUTMAN, F. J. & ROBERTS, F.M . (1962). Trans. Brit, mycol. Soc. 45, 449. NUTMAN, F.J.& ROBERTS, F.M . (1965). Kenya Coffee30 , 147.

86 7.2 Discussion

At the time that the results ofth e trialspresente d herewer e obtained,viz . 1964/1965,i twa s duly realized,tha t laboratory screening testsha d a limited value,especiall y as theobservation swer e onlybase d onth e inhibitiono f conidial germination in vitro (VERMEULEN, 1968).However ,considerin g theurgen t need to find atshor tnotic e fungicidesmor e effective thancoppe r compounds,i t was thendecide d tous e this relatively fasttechnique . Insubsequen t experimentsVIN E et al. (1973a)applie d for laboratory tests a screening techniquebase dbot ho n theabilit yo f fungicides todepres s sporulating capacity ofbar k andth e inhibition ofspor egermination ,thu sjudgin g the com­ bined effects ofth e two factors.Thi s screening technique ofVIN E and co-workers was thereforepartl ybase d onth eassumptio n that thedepressio no fth ebar k sporulating capacitywoul db e an indicationo fth eefficac y ofth e candidate com­ pound (NUTMAN& ROBERTS , 1970).Th e findings ofGIBB S (1969,1972) , GRIFFITHS et al. (1970a), HINDORF (1970,1972 ,1973a ,b ,c )an dVERMEULE N (1970a)drasticall y changed the concepto fsporulatin g capacity,thu sshiftin g theaim s ofdiseas e control from the suppressiono fth ebar k inoculum to thato f thecro p inoculum. Furthermore,i tshoul db ekep t inmin d that ingenera l laboratory screening of fungicides atbes toffer s only indications ofth e future disease controlling pros­ pects ofa candidat e compound. Considering theaforementione d changes inth e inoculumpotentia l concept,th e sporulating capacity datapresente d inthi s chapter shouldb e evaluateddifferent ­ ly, asth espra yapplicatio n ofth e selected compounds inth e fieldwa sbase d on thethe ncurren tpre-rai n recommendations (BOCK,1963 ;NUTMA N& ROBERTS , 1961). Thepreliminar y report,presente d at theFirs tSpecialis tMeetin g onCoffe e Research onEas tAfrica ,Nairobi ,statin g the favourable effects ofOrth oDifolata n onCB D incidence (VERMEULEN,1966b )attracte d little attention,a s this compound didno tmarkedl ydecreas e the sporulating capacity and thereforewa sno tcon ­ sidered tob e effective.Th e firmbelie f inth e inoculumpotentia l theory caused all favourable reports ona fungicide,whic hdi dno t suppress conidial production onbranches ,t ob e summarily dismissed.Th e same fatewa s allotted at the196 6 Meeting onCoffe eResearc h to the findings ofMULLE R (1964), reporting onth eCB D controlling effects ofcoppe r sprays,applie do ndevelopin g cropdurin g themono - modal rainy season inCameroon .Durin g asecon d visit toKeny aMILLE R (1968) recommended asimila r spray-schedule forKenya ,emphasizin g the importance of crop protectionb y fungicides,rathe r thanth e reductiono fth e conidialproductio n on thebark ,a sadvocate d inEas tAfrica .

87 InKeny a the copper spraysha d tob e appliedbefor e theonse to fth erain s inorde r toreduc e thepeak s of theS.C .o fth ebar k and thus decreasing thein ­ cidence ofCBD ,originatin g frombar k (BOCK,1963 ;NUBIA N &ROBERTS , 1961). Between 1956an d 1961 this spray schedule gaveadequat e results inhig h altitude coffeegrowin g areas inKeny a (NUTMAN& ROBERTS ,1961) ,bu tgraduall y evidence accumulatedtha tCB D could oftenno tb e controlledb y theapplicatio n ofcoppe r spraysbefor e the rains.Th e aforementioned change inconcep to fth e sporulating capacity andsubsequen t fieldtrial s led eventually toth e recommendation ofa continuous sprayprogramm e (GRIFFITHS &GIBBS ,1969 ;GRIFFITH S &WALLER , 1971 and GRIFFITHSe t al., 1971), similar to theCameroo n schedule (MULLER,1964 ,1968 , 1970, 1973,1978 ;MULLE R &GESTIN , 1967). Thequestio nho w copper,i nth e trialsbetwee n 1956-1961 , and copper and OrthoDifolata n inth e 1965 trials,applie dbefor e the rains,coul d stillhav e excercised amarkedly ,althoug hno t statistically significant,CB D suppressing effectha sbee n answeredb y GRIFFITHS andWALLE R (1971),VIN E andGRIFFITH S (1968) andVIN E etal . (1973a). Itwa s found thatOrth oDifolata n andmos t likely also copperwer e still active afterman yweek s ofexposur e toheav y rain,du et ore ­ distributionwithi nth e tree canopy (NEELY,1970 ,1971 ;PATEL ,1971 ;PEREIR A et al., 1973).A s thecompoun d thusprotecte d effectively the first stage? ofth e cropping cycle susceptible to C. ooffeanum, viz.flower s and expanding green berries (MOGK, 1970), thiswoul d explain the favourable anti-CBD effecto fOrth o Difolatan inth e fungicide screening trialpresente d inthi schapter . WhenVERMEULE N (1968)pointe d out theCB D controlling and other favourable effects ofOrth oDifolatan ,h e also stated that this compound didno t control leafrust , H. vastatr-ix. The application ofcoppe r fungicides to control the latter diseasewoul d stillb e required, also ifsubsequen t trialswoul d further confirm thevalu e ofOrth oDifolata n forCB D control.Late rexperiment s (VINEe t al., 1973a)hav e shown this tob e correct andOrth oDifolata n isstil l themai n anti-CBD fungicide inKeny a (FIRMAN &WALLER , 1977).

88 8. EFFECTS OF CULTURAL PRACTICES ON THE INCIDENCE OF COFFEE BERRY DISEASE ON TREES AND THE SUSCEPTIBILITY OF BERRIES

8.1 Introduction

Itwa sclaime dtha tcertai ncultura lpractice swoul d decreaseth eincidenc eo fCBD , especiallywhe nn ocoppe rcontainin g fungicidesha dbee napplie d (WIST& COLTMAN , 1966). Inorde rt oinvestigat eth eCB Dsuppressin geffect so fcultura lpractices , a serieso fobservation swer ecarrie dou ti n196 7-196 9o nthree ,adjacen t coffee estates (Table 12:A ,B an dC )i nth eKiamb uarea ,Eas to fth eRift ,altitud eap ­ proximately 170 0m an dcompare dwit h coffeetree sa tth eNationa lAgricultura l Laboratories,170 0m (D).Th edetail so fth eobservatio nsite sar epresente di n Table 12,usin gfo rth eassessmen to fvisua lCB Dincidenc eth esyste m listedi n 4.8: Theobservation s includedth efollowin g aspects: -th e C. coffeanum leveli nth ebar ko fsingl ean dmultipl e stemcoffe etree s grownrespectivel ywit hnorma lan dheav y fertilizersuse ,wit han dwithou tmulch , withan dwithou tsoi ltillage ,wit han dwithou tprunin gmodification san dwit h andwithou tcoppe rsprays ; -th esusceptibilit yo fexpanding ,non-coppe rsprayed ,gree nberrie st ostandar d conidialsuspension so f C. coffeanum ascompare dt oth esusceptibilit yo fex ­ panding,coppe rspraye dgree nberries ; -th epathogenicit yo f Colletotrichum conidialsuspension s frompruning san d mummifiedberrie s lefto nth egroun di ncoppe ran dnon-coppe r sprayedcoffe e fieldso nexpanding ,gree nberries .

This rangeo faspect swa sinvestigate db yusin gth esporulatin gcapacit y technique (25branches/site )an dth eassessmen to fth epathogenicit yo ngree n berrieso fth econidia lsuspension sobtaine d frombar ka suse di nth eepidemiologi ­ calobservation s (4.7).Al ltest swer ereplicate da tleas tsi xtimes .Conidia l sus­ pensionsobtaine db ywashin gbar kwer e standardizedo napproximatel y 2x10 conidia perml .

8.2Result s

Unsprayedcoffe eha da lowe r levelo f C. coffeanum, inhabitingth ecoffe ebark , thanspraye dcoffe ea salread yestablishe db yFURTAD O (1969,1970) .Heav yap ­ plicationso fchemica lfertilizer sha dn oeffec to nth eleve lo f C. coffeanum in thebark ,eithe ri nspraye do runspraye dcoffee .N oeffect so fhig hlevel so r 89 00,-N a oo

• 1-J •

i o c •rl CJ .3 U CO 3--- B •rl CU CU CJ -U O C co 9 « ki 3 O TJ U) cn TJ TJ CO CO eu cu 3 eu cu TJ eu cO T) rH 4J en •rl 3 I J! •H CJ 4H cu cj eu CJ > •H rl •n CO Xi x> TJ Xi Xi Tt 00 4J •r l 0) >f00t r-* cu •H 3 Vi cu cn rl eu 3 en CJ i—i 3 CO Tl cu -H -.H TJ eu •rl -rl u •H CJ •H x: i-i u cu .3 rH TJ rl (1) Xi TJ >o TJ a xa > a eu Xi T! rl Tl 0) Fi CD o P. cu o n a. eu eu 3 eu G

CO VJ CO rl u eu cj a) T) a) CU 3 N B B -H T3 B "rl -O CU •rl g -a B rH 1— 1—1 CO s (-] •ri SH CO 42 4J o 4-1 01) rl a c) 14 C) o eu cu a. CD CD rl Z MH Pi C4H P. TJ I eu 5 rH s I J3 I oj CU Xi co CO CJ D T3 CM Tj 00 U W C ftD O co TJ 3 OO'H CU 00, TJ en i en -j en •" cO cO B Xi CO rl 3 rH 00 Cn J*i XI rH 00 • o 3 c co 3 co i—i CO O ej 3 *H en O rH O CJ rH rH CO 3 S CO U CO ,3 -,-1 3 -H rl 3 CU -rl eu 3 o ri O 3 4J CUHH CU 3 -rl cu MX! CJ 4J •rl 3 X2 J3 B 4-1 Bi E en p, B -rl en CO O Pi S u 3 W

xi -a TJ o #i 3 CO MH -H 00 • CJ i-l CJ •rl CJ •rl CJ •rl rl ej •rl 3 3 en 3 en 3 en CO 3 cn CJ i-l XI• en i-31-3 cu en J • •J eu en i-3 r-•l u eu cn u cO rl •rl C/3 C/l Pn S m c«' a : tn g

en 1 • ft H 3 en CO CO TJ en • •• cn R O en o >*B > 3 .» 3 cu 3 B 3 O c 4J E -H cu O 1 eau B en 00 1 csu O 1 cu o CJ •r()l CO eu 4-1 B 4J •H cu en B 4-1 cu CU 3 eu CX B 4J •rl I CO en ew ^H •rl 3 4J rH H 3 3 CU i-l •rl CU SO 14-1 •H CO cu 00 .-H en 4-1 - paq 00 cu 00 M-l •rl CO CU 00 UH •H o C/l 00 "0 i—i 3 •i- i 3 TJ r^ 3 00 3 oo c_> 3 i—i 3 •rl 3 •3 rH 3 -rl 3 •H XI i—l 3 Q P. •r* -a 3 3 a •rl 3 O CU 3 •H o. •rl •3 3 3 a •rl TJ 3 4-1 o CO •i- i en o iJ CU •rl en •rl in rl •H eu cn •rl en O !H CU •H en 0 rl cfl U •H3 a 4-1 B a, E 4J en 0 CU 3 -a en 4J B P. B 4J H P.-a eu rl O u 3 3 0 3 o 3 3 O 3 cu 4-1 i—t PH eo cd 3 •rl CJ S CO o IH cn a. 4J p. S3 CO 3 •rl o S cO 3 •H 3 s a 3 O •i- l <) ri 3 fill en ^-\ 1 CU 3 0 CO r-l-r! CU en CO •rl < 3 TJ ri CO 4-> X! ctl Tl C>O^ 4-J i—i o en •i- i cu u a o cu 3 00 CD SH CO CO a a) P. 3 cn • rl V.C3 TJ p. S3 3 ri C>U en 0 •H CT\ 3 cn cu u O cu 3 i -a TJ 3 Tt rH r— 0) rH o •H co en en eu n CU O f-j rl 3 <4H 4J CO cO i^-rl CU -rl •" ^-t 1-1 )-i CU TJ s CO ai cu CO rH P. rH ^D CO 4J r^ft CT \ O Cu 0) H Z en g n a. P- P. \D CJ 3 vO ^D CJ p,Xi p. a o aen •H C> CA CTi en co CJ co — B a —r "

CO H

90 organicmaterial ,includin gmulching ,an dvariation s inprunin g systems onth e C. ooffeanum level inbark ,coul db e established.Apar t fromth e aforementioned differences between sprayed andunspraye d coffee,th e othercultura l measures haven oothe reffec t than improving thegenera l conditiono fth e coffee trees. The CBD susceptibility ofgree nberrie s obtained fromunspraye d coffee treesprove d tob e lower after inoculationwit h standard suspensions of C. ooffeanum, thanth e susceptibility ofgree nberrie s from routinely sprayed trees (Table 13). Prunings andberrie s lefto n the ground for some timedi dno tyiel d any conidial Colletotviohum suspensions,capabl e tocaus e infectiono ngree nberries .

Table 13:CB D infection percentages of green berries of sprayed and unsprayed coffee trees,inoculate d with a standard suspension of C. ooffeanum conidia of 2x10^ conidia perml .

1st 2nd 3rd 4th 5th 6th test test test test test test

Berries from unsprayed 45 50 46 42 36 35 coffee trees Berries from sprayed 70 82 68 64 61 57 coffee trees

8.3 Discussion

Apart from thealread y available evidence,tha t copper sprayed coffee treeshar ­ bour ahighe r level of C. ooffeanum inth ebark ,n o substantial indications have thusbee n foundtha tothe rcultura lpractice smigh t affect the C. ooffeanum level incoffe ebar k toan ypractica l oreconomica l extent.Allege d instances of lower CBD incidence oncoffe e trees,treate d onlywit hhig h levels oforgani cmateria l orspecia l cultural practices asdescribe d inTabl e 12,hav eno tbee nsubstanti ­ ated inthes eobservations . Itseem s that the effects of fungicidal applications have beeno fgreate r importancewit h regard toth e C. ooffeanum level inth e coffeebark ,tha n factors like organicmanurin g or culturalpractices .STEINE R (1972a,1973d ) found that the application offolia r fertilizers in combination with fungicides had an increasedCB D controlling effect,a scompare d to fungicide sprays alone.NUTMA Nan dROBERT S (1969e)o nth econtrar y claimed ahighe r levelo f CBD afterapplicatio no f foliar fertilizers,thu s challenging the results of DAPONT E (1966)indicatin g that foliar applications ofcalciu m super phosphates controlled CBD inAngola .Th ewhol ematte r ofth e foliarapplicatio n of

91 fertilizers shouldb e further investigated, especially inrelatio n to their possible effectso n thephyllospher e ofcoffe ebranche s (GIBBS,1972 ;ROBINSO N& WALLER, 1966;WALLER , 1972a). With regard to the lowerCB Dsusceptibilit y ofgree nberrie s fromunspraye d coffee trees (VERMEULEN, 1970c)th e following explanation canb e offered.ADUAY I (1972)reporte d in hiswor k onth e effects ofcoppe r sprays onth eminera l nutrient content ofcoffe e seedlings.HINDOR F (1973d)establishe d alsotha tyoun g leaveso n regularly copper sprayedplots ,bu t developed after the last sprayap ­ plication,containe d six timesmor e copper than identical leaves onnon-coppe r sprayed sites.A similar increase ofcoppe r content inth edevelopin g berries of sprayed coffeemigh t thusb e expected and these levels ofcoppe rma y have hada stimulating effecto nth e germination of C. coffeanwn (NUTMAN& ROBERTS , 1962b;RICHARDSO N Þ , 1962). This aspectneed s tob e investigated inmor e detail inorde r toestablis hwhethe r orno t only copper fungicides tendt o induce ahighe r level ofCB D susceptibility indevelopin gberries .

92 9. GENERAL DISCUSSION AND CONCLUSIONS

Researchpresente di nthi spublicatio n constitutesa par to fth eentir e research efforts since196 4directe da tcoffe eberr y disease {Colletotriahum coffeanum), themajo rcoffe eproble mi nKenya .Th eresult sobtaine dar etherefor ediscusse di n conjunctionwit ha nevaluatio no frelevan t researcho fothe rworker sa ttha ttime . Fromth eliteratur e reviewi nChapte r 3an dals o fromth einformatio npre ­ sentedi nth eothe rchapters ,i tbecome s cleartha tth eresearc h findingsob ­ tainedi nth emid-1950' s haveha da considerabl e impacto nth eresearc h carried outi nth eperio d 1961-1968 . The adherencet oth e195 6-196 1 findingswa sbase do nth efir mbelief ,tha t thesprea do fCB Dint oth elo waltitud ecoffe e areaswa scause db ya chang ei n climate,mainl yb ya nincreas e inrainfal l (NUTMAN& ROBERTS ,1969d ;WALLIS , 1964), buttha totherwis en oessentia lchange sha doccurred . Inth eyear s 1964-196 8th e policyo fth eCoffe eResearc hStatio ntoward sCB Dremaine d therefore foundedo n theol dcontro l concepts,i.e .applicatio no fpre-rai ncoppe r sprays,whichwer e then stillaccepte da svali dan deffectiv eb yth efarmer san dth eresponsibl e government officials.Mos to fth eCB Dresearc hi nth eearl yan dmid-1960' sha st ob econ ­ sideredan devaluate d against this background. The international coffee contacts improved afterth eFAO-Technica lworkin g partyo ncoffe eproductio nan dprotectio ni nBrazi l in1965 .Coffe eauthoritie si n Kenyabecam e awareo fth efact ,ho wlittl ewa sknow nabou tth eCB Dpathogen .I n additioni tbecam eknow n thatth ediseas ewa sno tonl y confinedt oEas tAfric aan d theneighbourin g Zaire (Belgian Congo),bu twa sals opresen ti nAngol aan d Cameroon.Thu sCB Dwa slifte dou to fth erathe r limited- an dexclusiv e- Kenya n contextan dinternationa l helpan dknow-ho w couldbrin g abouta break-throug ho f theol dCB Dconcept san da ndeepene d insighto fCB Dresearc haspect s inonl ya fe w years. The resultspresente di nChapte r 5illustrat eho wth equestion s aroundth e compositiono fth e Colletotrichvm complexan dth eepidemiologica l roleo f G. aingulata were jointly answereda tth een do fth e1960's ,largel ya sa resul t ofth eresearc hactivitie s ofFURTAD O (1969,1970) , GIBBS (1969),HINDOR F (1970, 1973)an dVERMEULE N (1970a,b) .I tremain sunclear ,however ,why n olin kcoul db e foundbetwee n C. coffeanum and G. aingulata inKenya ,whil e thisrelationshi pha d been reported fromTanzani a (HOCKINGe tal. , 1967;JOHANNS ,cite db yHINDORF ,

93 1972).Withou t anydoub t it isnecessar y to initiate geneticwor kwit h Colle­ totrichum isolates ofcoffe e inorde r toestablis h theirpathogeni cvariabilit y in relationt oheterocaryosi s (STEPHAN,1968 ;STROBAL , 1960).Th ewor ko fWHEELE R (1954)o nth egenetic s of G. cingulata shouldb e incorporated inthi sne w research programme. The research findings during the late 1960'sals o ledt oa completel y differ­ ent evaluationo fth e 'inoculumpotential ' theory ofNUTMA Nan dROBERTS ,whic h had hithertobee ngenerall y accepted.Th e inadequacy of theusag e ofth e totalsporu - lationcapacit y dataa sa.mean so fpredictin g disease incidence,i s illustrated by thepathogenicit y testscarrie d outwit h conidiaobtaine d from individual inter- nodes,a spresente d inChapte r 6.Th e actual rhythmo fconidia lproductio no f C. coffeanum and theperiod s ofpathogenicit y ofthes econidi a frombar k arede ­ finedbot h intim e (during orafte r the rains)an d location (internodeswit h the firstsig n of maturation)b y the testso ngree nberrie swit h theconidia l sus­ pensions frombark . Inconjunctio nwit h thedetaile d studies ofWALLE R (1972a)thes e findings provide an indication ofth e timewhe n thebar k inoculumwil lb e available toin ­ fectth e susceptible stages ofth e crop (MOGK,1970 ;I^DG K& HINDORF , 1975).Con ­ sidering thepeak s inpathogenicit y of thebar kpopulatio n of C. coffeanum and the findings ofWALLE R itmay thenb econclude d thatth e shift inS.C .compositio n of thebar k towards C. coffeanum will occuri nth emiddl e ofth erain y season,whe nth e very susceptible expanding greenberrie s arepresen t onth ebranches .Onc eberr y inoculum isavailabl e itwil l quickly surpass thebar k inoculum,especiall y when sufficient susceptible crop ispresen t (GIBBS,1969 ;GRIFFITH S &WALLER ,1971 ; GRIFFITHS et al., 1971;MJLLNGE ,1971a ; WALLER, 1972a).Th e reportedhighe rin ­ cidence of G. c-Lngulata oncoppe r sprayed coffee trees (NUTMAN &ROBERTS ,1969c ; VERMEULEN, 1970b)seem s tob econtradictor y toth e shift ofth e Colletotrichum population towards C. coffeanum. Most likely the increase of G. cingulata might be comparedwit h the influence ofdecreasin g temperatures onth e initiationo fper ­ fectstage s ofothe r fungi inth etemperat e climate.Moreover ,peritheci a of G. cingulata haveneve rbee n foundo n the restrictedbar k area,occupie db y C. coffeanum. The subjecto fbar k colonizationb y Colletotrichum spp.need s tob ein ­ vestigatedmor e closely.Ou rknowledg e ofth e reasonswhy coffeebar k becomes colonized, especiallyb y theCB D fungus isstil lver y limited and it isals oun ­ knownwh y Colletotrichum species occupy strictly definedbar k areasonly .Evi ­ dence that C. coffeanum would re-infectbar k fromberr y inoculum (GIBBS,1972 ; NUTMAN &ROBERTS ,1969d )i s insufficiently documented.

94 Experimental evidencepertainin g to theepidemiolog y ofCB D throughKenya , theadjacen t EastAfrica ncountrie s andultimatel y Ethiopia,Zaire ,Cameroo n andAngol a isals o limited. ForKenya ,Ugand aan dTanzania ,th egenera l accepted theory (NUTMAN& ROBERTS ,1961 )wa s that thediseas ewa s transported asbar kin ­ oculum inth ebranche s ofseedlings .Thi smigh thav ebee ntru e ina numbe r of cases inKenya ,Ugand ao rTanzania ,bu t itcanno t fullyexplai nal lCB D outbreaks inEas tAfrica . Inparticula r the theory doesno t account forth e outbreaks in Angola,Cameroo nan deve nEthiopia ,a s itconsider s EastAfric a the original centre of infection. The resultso fexperiment s inAfrica n smallholder coffee areas (Chapter6) , were inagreemen twit h those obtained inth earea swhic hwer epreviousl y research­ edb yNUTMA N andROBERT S (1961).Accountin g forth e different rainfall inMeru , where themai nprecipitatio n issi xmonth sbehin d thato fth eothe r areas studied, all resultswer e similar. The applications offungicide s during the rains in196 9ha d amarkedl y dis­ easecontrollin g effect,a spredicte d forKeny ab yMULLE R (1968)an dalread yin ­ dicatedb y theKitale-Mer uspra y timing trials.Th e change frompre-rai n sprays tosprayin g during the rainswa s slow inKenya . Ittoo k theCR S some time to realize that thepre-rai n spraying schedulewa s toounreliabl e and largelyin ­ effective (GRIFFITHS& WALLER , 1971),befor e spraying during the rains became theirpolic y (GRIFFITHS et al., 1971).Th e incentive for this change inth e spray programmewa sundoubtedl yprovide db yMULLE R (1964,1968) . One ofth e first fungicides,whic h lookedpromisin g inscreenin g trials,wa s OrthoDifolatan ; itseffectivenes s inCB D control has since thenbee nconfirme d in additional trials.Th e compound isstil lwidel y applied afterbein guse d already fora perio do f sometwelv eyear sdespit e the introduction ofsystemi c com­ pounds sucha sBenlat e (VINE et al.,1973a) .Th e decreasing effectiveness of Benlate and related compounds,becaus e ofth ebuild-u p offunga l resistance (OKIOGA, 1976),hasdiminishe d thediseas e controlling prospects of these systemic fungicides.Ultimatel y futureCB D controlwill ,however ,mos t likely rely more on systemic rather thannon-systemi c compounds,i nlin ewit h thenatur e of thebar k colonizationb y C. coffeanum. Copper fungicideswil l remainpar t ofth e spraying schedule,a sthe yar estil l themos t effective compounds for leafrus t control,unles s specific rust controlling fungicideswil lbecom e available. GRIFFITHS (1972)alread ypointe d outth eurgen tnee d tokno wmor e about the inter­ actionsbetwee n fungicides,coffe ean d thepathogen s oncoffee ,especiall y because ofth e increased applications ofanti-lea f rust fungicides inth emajo r coffee producing countries (WALLER, 1972b).Considerin g the experience inKeny awit h

95 fungicides ingenera l andth eresult spresente d inthi spublication ,th eurgenc y to investigate theaforementione d interactions isstil lpresent . The alleged CBDsuppressin g effectso fcultura lmethod shav ealread ybee n discussedb yRAYNE R (1952).Th e resultso fth e comprehensive observations,pre ­ sented inthi spublication ,d ono t showan ydifferenc e indiseas e incidencebe ­ tweena wid erang eo freputedl y effectivemeasures .Th e alreadymentione dhighe r bark incidenceo f C. aoffeanum incoffe espraye dwit hcoppe r fungicides isagai n evident,regardles s ofth ecultura lpractic e applied. The increased susceptibilityo fcoppe r sprayedberrie s suggests,tha ti nth e years 1956-1961 ,whe npre-rai ncoppe r sprayswer e successfully applied,th e quantities ofcoppe ro nan d incoffe e tissueswoul d declinegraduall ydurin g the rainsunti l apoin twher e theremainin g copper levelwoul d actually favour thein ­ fectionb y 0. aoffeanum (NUTMAN& ROBERTS ,1962b ;RICHARDSO N Þ , 1962).A s thiswoul dhappe ndurin g themos t susceptible stages ofth e cropping cycle (MDGK, 1970)i tmigh texplai nth econsiderabl e losses incurreddurin g theperio d 1962- 1966whe nth edistributio no fth erainfal l differed from that in195 6-1961 , thus affecting theredistributio nan d levelo fdeposit s ofcoppe r fungicides inth e coffee trees (GRIFFITHS& GIBBS ,1969 ;GRIFFITH S &WALLER , 1971). Inrelatio nt oth eaforementione d aspectso fCB D research finally theim ­ portant findingso fSTEINE R (1973b)wit h regard toth e relationshipbetwee nrain ­ fallquantitie s andne wCB D infections provided the firstexperimenta l evidence about theoptima l applicationo f fungicidal sprays.Als o related tothes easpect s was thewor k ofGASSER T (1976,1978 )i nEthiopia ,indicatin g thegrea thuma ninter ­ ference inth eCB Depidemiolog y andth e favourablediseas econtrollin g prospects of 'cleanpicking' ,viz .removin g allaffecte dberrie sb yhan d atregula rintervals , thuspreventin g theescalatio no f theCB Doutbreak s through 'berry-to-berry'in ­ fection (GIBBS,1969 ;GRIFFITH S &WALLER , 1971;GRIFFITH Se t al., 1971). The researchdat apresente d hereprovid e informationo nth e following aspects ofCBD : a.Th e locationan d frequencyo foccurrenc e ofsom ecomponent s ofth e Colletot- riahum populationo nth ebar k inrelatio n tofungicid eapplications . b.Th e roleo f G. cingulata inth ediseas e epidemiology andth e absenceo fa relationshipwit h C. aoffeanum hasbee nestablished . c.Th e sporulating capacityo fbearin g branches interm s ofpathogenicit y on greenberrie s asa functio no frainfal l andfungicid eapplication . d.Th epathogenicit y ofth ebar k inoculumwhic h increasesdurin g therains , largelybecause : - thenumbe ro fsaprophyti c Colletotviohum conidiadecreases ,als odu et opre -

96 rainfungicid eapplications ; -th e C. coffeanum containing tissues increasewit hth egrowt h flushesa tth e onseto frains ; -th ebar k surfaceo fth einternode swit hth efirs t signso fbar kmaturatio n offers lessopportunit y thano nolde r internodesfo rfungicide st oremai n attached,thu sprovidin gth efungu swit h suitable sporulationsurfaces . e.Th eeffectivenes so fOrth oDifolata na sa nanti-CB D fungicide. f.Th ehighe r susceptibilityo fberrie s sprayedwit h copper fungicidesa scompare d withnon-coppe r sprayedberries . g.Th eineffectivenes so fcultura lmethod si nCB Dcontrol . The following topicsar esuggeste da sprioritie s infutur eresearch : -a world-wid e classificationo f Colletotrichum spp.,inhabitin gth ebar ko f arabicacoffe ei norde rt opredic tpossibl e outbreakso f C. coffeanum; -continue d investigations intoth eeffect so fcoppe ran dothe r fungicideso nth e compositiono fth e Colletotrichum barkpopulations ; -detaile d studieso fth eanatomica l preferenceo f C. coffeanum forcertai nbar k areaso fcoffe etwigs ; -a stud yo fth egenetic so f Colletotrichum spp.an dthei rpossibl e relationship with G. cingulata; -th eselectio no feffectiv e fungicides,wit h special referencet osystemi ccom ­ poundsan dth ebuild-u po ffunga l resistancet othes e fungicides; -th econtinuatio no fbreedin g forCB Dresistance . Althoughth eCB Dproble m inKeny ai scertainl y lesscalamitou sno wtha ni t wasi nth emid-1960's ,i tstil l requires continuous attention;th edevelopmen to f more resistant coffeecultivar s seemso fparticula r importance.Th esam eapplie s toothe rAfrica ncoffe ecountries ,wher eCB Dinflict scro plosses . Since H. vastatrix wasfoun di nBrazi l in1970 ,rigi dcontro lmeasure sar e carriedou ti nth eSout h andCentra lAmerica ncoffe egrowin g countries (WALLER, 1972b;WELLMAN , 1970).Considerin g theeffect si nKeny ao ffungicida l spraysi n favouro f V. coffeanum, careful attention shouldb egive ni nth eaforementione d countriest osymptom s similart othos e causedb yCB Di nKenya .Th ereporte d field incidenceo fCB Di nBrazi l (FIGUEIREDO,persona lcommunicatio nt oDr .D .MULDER) , isalarmin ga sthi swoul d endangervas t areaso fcoffe eno tonl yi nBrazil ,bu t alsoi nth eothe rSout han dCentra lAmerica ncountries .Th eassume d inabilityo f CBDt osprea dquickl yan dinflic theav y lossesunde rth eclimati c conditionsi n thosecountries ,shoul db eviewe dwit h great caution,a sespeciall yth eclimati c andcultura l conditions inth emontan e areaso fCentra lAmeric amigh tb econ ­ ductivefo rCB Doutbreaks . International cooperationno tonl yi nth efiel do f

97 leafrus teradication ,bu tals o incontrollin g coffeeberr ydiseas e isnecessar y andrequire sth egatherin g ofdetaile d informationo nth e Colletotr-Lchwn bark colonization fromal larabic acoffe e growing countries inth eworld .Thi s co­ operationshoul dals o includea programm e fortestin go fCB D resistantmaterial , as initiated inKenya .

98 10. SAMENVATTING

Doorhe tontbreke nva nnatuurlijk egrondstoffe ni sd eeconomi eva nKeni avoo ree n grootdee lafhankelij kva nd elandbouw .Uitgestrekt e gebieden zijnbeplan tme t arabicakoffie , Coffea arabioa, end eexpor tva nkwaliteitskoffi ei see nbelang - rijkebro nva ninkomsten . Vooraln ad e2 eWereldoorlo g ishe taandee lva nd eklein eboe ri nd ekoffie - cultuurtoegenomen .Tegenwoordi gi sbijn ad egehel ecultuu ri nhande nva n Afrikaanse landbouwerse n-ondernemer se ni sd eeertijds e dominantieva nEuropes e koffieplanterse nfirma' sverdwenen . Deontwikkelin gva nd ekoffieteel ti nKeni ai s- behalv edoo rd eeconomisch e terugslag ind ejare nderti ge nd e2 eWereldoorlo g -oo kbelemmer ddoo rhe top - tredenva ntwe eschimmelziekten :d ekoffieroest ,veroorzaak t door Hemileia vastatrix end ekoffiebesziekte ,veroorzaak t door Colletotv-iohim aoffeanvm. Deze laatste ziekte,waarbi jbloeme ne nbesse ni nd everschillend eontwikkelingsstadi a wordenaangetast ,i she tonderzoe kva ndez epublikatie ,waari ngepubliceerd ee nno g ongepubliceerde onderzoekgegevensui td eperiod e 1964-196 9worde ngepresenteerd . Depublikati ebegin tme tee nkor toverzich tva nhe tgehel eonderzoe kverrich t sindsd eeerst ewaarnemin gva nd ekoffiebesziekt e inWes t Keniai n1922 .I nd e periode 1922-195 0wer dvastgesteld ,da td eschimme l C. ooffeanum opgroen ekoffie - bessenziektesymptome nveroorzaakt .I nexperimente nbleke nfungicide nnie to f weinig effectieft ezijn .Oo kd eproeve nme tkunstmes ttoonde naan ,da tverhoogd e kunstmestgiftengee n invloedhadde no phe tvoorkome nva nd eziekte .Slecht she t gebruikva nziekte-resistent e rassenlee kron d195 0perspectieve nt ebieden . Gedurended eperiod e 1950-196 6wer dmee r inzichtverkrege ni nenkel eas - pectenva nd eziekte .Z ower dvastgesteld ,da the tpathogee ni nd ebas tva nkoffie - takkenleef te nva ndaarui td einfecti eva nbesse nveroorzaakt .Toe nbespuitinge n metkoperhoudend e fungicidenvoo rhe tbegi nva nd eregen sgunstig e resultatenop - leverden,naa rme naanna mdoo ree nverminderin gva nhe tbastinoculum ,lee kd e ziekteblijven dt ekunne nworde nonderdrukt . Hetoptrede nva nd ekoffiebesziekt ewa saanvankelij kbeperk tto td ehoogge - legenkoffiegebiede n (boven168 0m )va nWest -e nMidde nKenia .N a196 2kwa md e ziekte echteri ntoenemend emat evoo ri nlage rgelege ngebieden .Aa nhe tein dva n de zestigerjare nware nall ekoffiegebiede ni nKeni aaangetast .D ehierdoo rver - oorzaakteverlieze nware nbuitengewoo nhoog ,hetgee nerto e leidde,da ti n196 6he t

99 onderzoeko phe tterrei nva nd ekoffiebesziekt ewer dgeintensiveerd .Hierdoo rwer d ooko pinternationaa lnivea usteu ngezoch te ni n196 7werde nonderzoeker se nfonds - enbeschikbaa rgestel ddoo rd eregeringe nva nEngeland ,Wes tDuitslan de nNederland . Insamenwerkin gme tKeniaans eonderzoeker swer d alsgevol ghierva nsne ln S 1966ee ngrot evooruitgan g gemaaktme tbetrekkin g totd ekenni sva nd eepidemiolo ­ gic,biologi ee nbestrijdin gva nd e ziekte.D enieuw e inzichtenmaakte nhe to.a . mogelijkd eziekt eo peconomisc hverantwoord ewijz echemisc ht ebestrijden .D egun - stigeresultate n inKameroe nhebbe ndaarbi jhe tonderzoe k inbelangrijk emat e gestimuleerd. Dei ndez epublikati egepresenteerd e resultatenbetreffe nenerzijd sreed sge - publiceerdegegevens ,anderzijd see nvri jgrot ehoeveelhei d informatie,di eno g nieteerde rwa sgepubliceerd . Inhe thoofdstu kove rd eanatomisch ee nmycologisch e aspectenword t ingegaano phe tvoorkome nva nhe tpathogee n ind ecorte xva nd e bastva nkoffietakke nte ntijd eva nd evormin gva nd eeerst ekurklaa gdaarin . Detakgedeelte nwaa r ind ecorte xreed smee rkurklage n zijngevormd ,bevatte nal - leensaprofytisch e Colletotvichum soorten. Glomerella cingulata, hetperfect e stadiumva nd esaprofytisch e Colletotvichum soorten,speel t inKeni awaarschijn - lijkgee nro li nd eepidemiologi eva nd ekoffiebesziekte .Daarentege nzij ne raan - wijzingen,da td eschimme li nTanzani ava nhe t Glomerella stadium in C. coffeanum kanovergaan . Voortswer d aangetoond,da ti nKeni amee r C. ooffeanum voorkomto pkoffie - bomen,di eregelmati gworde nbespote nme tkoperhoudend e fungicidenda no ponbe - spotenbomen .Di twijs to pee nverschuivin g ind e Colletotrichum-Tpopulatie ten gunsteva n C. coffeanum onder invloedva nd efungicide nbehandeling . Eveneenswer dgevonde nda to pme tkopermiddele nbespote nkoffietakke nd e vruchtlichamenva n G. cingulata invee lgroter eaantalle nvoorkwame nda no ptakke n nietbespote nme t fungiciden.He tverhoogd e aantalperitheci ava n G. cingulata is echternooi taangetroffe no pd ebastgedeelte nme td eeerst etekene nva nbastrijping , waarinhe tpathogee nva nd ekoffiebesziekte , C. coffeanum,voorkomt. Deepidemiologisch egegeven swerde nverzamel di ntwe egebiede nme tuit - sluitendklein ekoffietuine nva nAfrikaans eboeren .D ewaarneminge nomvatte nd e sporulatie-capaciteitva nhe t Colletotrichum-camplex vand ebas tva ntakke ni nre - latieto td etijd ,d eregenva le nhe tgebrui kva nfungiciden .Voort swer dnagegaa n hoed esporulatie-capacitei tva nd ekoffietakke nsamenhang tme td epathogenitei t vandi tbastinoculu mo pgroen ebessen . Uitdez egegeven sblee kopnieu wda t C. coffeanum alleenvoorkom ti nd e jongere takgedeeltenwaa rd erijpin gva nd ebas tpa s isingezet .Slecht sd e conidienverkrege nva ndez etakgedeelte nkonde nn a incubatieonde rvochtig eom -

100 standigheden infectieso pgroen ebesse nveroorzaken .Daarnaas twer dwee rvastge - steld,da td epopulati eva nd eparasiet , C. ooffeanum, duidelijk groterwa s inme t kopermiddelenbespote nvelde nda n inonbespote nkoffie .He tmoment ,waaro pbastin - oculumva nd e ziekte aanwezig is,val t samenme the tmees tvatbar e stadium ind e vruchtvorming.Di t ishe t stadiumwaari nd ekoffiebe s sneluitgroei t enwaarbi j de klimatologischeomstandigheden ,t.w .hog e relatievevochtighei d engunstig ekiem - en infectie-temperaturen,optimaa l zijnvoo r infectie.Al skoffiebesse nzij nge - infecteerd,word t deuitbreidin g vand e ziekte spoedigversnel ddoo ree ncontinue , hogeprodukti eva nconidie no pbessen . Inhe thoofdstu k overd ekeuz eva ngeschikt e fungicidene nhe tbepale nva n het juiste tijdstipvoo rd e toepassing ervanword t ingegaan oponderwerpen ,di e in de zestigerjare n inKeni a zeercontroversiee lwaren .Enerzijd swa sme n indi e tijd werkelijk overtuigdva nhe tgunstig e effectva nbespuitinge nme t kopermiddelen toegepastvoo r de regens.Anderzijd swerde nsteed swee r stemmengehoord ,di enie t alleend ejuisthei dva nhe t tijdstipva nspuite n intwijfe l trokken,maa roo kd e effectiviteitva ndez e fungicidenvoo r debestrijdin gva nd ekoffiebesziekt ebe - twijfelden. Inhe t raamva ndez e ambivalente situatiemoe td eopze tva ndi tfun - gicidenonderzoekworde ngezien .Gedurend ed euitvoerin gva ndez eproeve nblee k reeds spoedig datd egangbar egehanteerd e inoculum-potentiaaltheorie end econ - sequenties,di eui td e inoculum-potentiaalgegevenswerde ngetrokke nme t betrekking totd everlieze nveroorzaak t doord ekoffiebesziekte ,nie twerde nondersteun d door de ziektewaarnemingeni nhe tveld .Uitgaand enamelij kva nd e inoculum-potentiaal- resultatenwer dhe tmidde lOrth oDifolata ndoo rander eonderzoeker s aanvankelijk nietbeschouw d alsee neffectie f fungicide,alhoewe l deresultate n inhe tvel d veelbelovendwaren .D euiteindelijk e resultaten,welk eme tOrth oDifolata nwerde n bereikt,ware nhoopgevend ,hetgee nvoora lwer dbevestig d bij latere toepassingen gedurende deregentijd .Di tmidde l istegenwoordi g nog steeds Senva nd ebelang - rijkste fungicidenvoo rd ebestrijdin g vand ekoffiebesziekt e inKenia . Wathe t tijdstipva ntoepassin g vanbespuitinge nbetreft ,beweze nd eproeve n dejuisthei dva nd eresultate neerde rbereik t inKameroen ,waa rd e fungicidennie t werdentoegepas t terverminderin gva nhe t 'bastinoculum',maa rvoo rhe t aanbrengen van eenbeschermend elaa g rondd evruchten . Deveronderstelling ,da tbepaald e cultuurmethoden eengunstig e invloed zouden hebben ophe toptrede nva nd ekoffiebesziekte,ko nnie tworde nbevestig d doorwaar - nemingen inhe tvel d en inhe t laboratorium. Invelde nme thog e giftenkunstmest , stalmest,organisch e stofen/o f intensieve snoeisystemenwer d geen lagernivea u vand e ziekte aangetroffen.D e eerdergesignaleerd everschille n inhe t C. ooffeanum- niveau tradenallee no p tussenbespote ne nonbespote nkoffie .Voort sbleke nbesse n

101 vanonbespote nkoffiebome nminde rvatbaa r te zijnda nbesse nva nbespote nbomen . Inaansluitin ghiero p dienthe twer kva nee nDuits e onderzoeker inKeni at e wordengenoemd ,waardoo rko nworde nvastgestel d nawelk ehoeveelhei d regenopnieu w moetworde ngespote nme t fungicideno mnieuw euitbarstinge nva nd e koffiebesziekte tevoorkomen . InEthiopi ebewee see nander eDuits eonderzoeker ,da td emen s een belangrijkero lspeel tbi jd everspreidin g vand eziekte .Voort swer d aangetoond, dathe t 'schoonplukken'va nd ebomen ,d.w.z .he t regelmatigverwijdere nva naange - tastekoffiebessen ,ee nbelangrijk e factorwa s ind e ziektebestrijding.Dez eonder - zoekgegevens zouden,mee rda nvoorheen ,diene nt eworde n ingepast inee nbree dop - gezetprogramm avoo rd eeffectiev ebestrijdin g vand ekoffiebesziekte ,nie t alleen inKeni amaa roo k inander ekoffieverbouwend e landen. Indez epublikati eword t enkelekere ngeweze n ophe tverschuive nva nd e Colletotrichum-balans inKeni a ind erichtin gva n C. coffecmwn tengevolgeva n langdurig spuitenme t fungiciden. InZuid - enMidde nAmerika ,waa rme nsind s 1970 dekoffieroest , H. vastatrix, bestrijdtme tkoperhoudend emiddelen , zouee nder - gelijkeverschuivin g ind e Colletotrichnm-populatie eveneenskunne noptreden . Recenteberichte nui tBrazili ewijze ne r inderdaad op,da t C. ooffeanum daari n hetvel do pkoffiebesse n schijnt te zijnaangetroffen . Wathe t toekomstigeonderzoe kbetref tword t aangedrongeno pee ninternational e samenwerkingtusse nkoffieproducerend e landeno phe t terreinva nd e inventarisatie vanbastbewonend e Colletotriahwn spp.o m zodoende tekunne nvaststelle n of C. ooffeanum aanwezig ise ndu spotentiee l eengevaa rvormt .Daarnaas t zalhe t toetsenva nhoogwaardig e cultivars opresistenti e tegend ekoffiebesziekt e ook internationaalmoete nworde nvoortgezet .

102 11.SUMMAR Y

Dataar epresente do nresearc hi nKeny ai n196 4- 196 9o nanatomical ,mycological , epidemiological,chemica l controlan dcultura l aspectso fcoffe eberr ydisease , Colletotrichum ooffeanvm Noack,o f Coffea arahica L.Th epathoge ncause sflowe r andberr y lossesan dwa sfoun di nbranche swher e itoccupie d clearly defined areas ofth ecorte xjus tbefor eo rafte r formationo fth efirs tphellogen . Saprophytic Colletotrichum spp.inhabi tbar k areaswit hmor eperiderm s inth ecortex .N o relationship couldb efoun di nKeny abetwee n Glomerella cingulata (Stonem.)Sp .& Schr.,th eperfec t stageo fmos to fth esaprophyti c Colletotrichum barkcomponents , and C. coffeannm. Theseasona l fluctuations inpathogenicit y inth ebar kpopula ­ tiono f C. coffeanum couldb eassesse dan dcompare dwit hth etota l sporulating capacityo fth ebar kpopulatio no fal l Colletotrichum spp.Formerl y theleve lo f this total sporulating capacity,o r 'inoculumpotential 'a si twa sthe ncalled , wasuse da sa nindicatio nwhe npre-rai n copper spraysha dt ob eapplie dan dho w effectively thefungicid eha dreduce d thebar k inoculum.Base do nthes e datath e recommendations forchemica l controlwer e changed frompre-rai n fungicideappli ­ cations,t oa sprayin g regimewel l intoth erain yperiod ,th eaccen tbein go n protectiono fth eberrie s rather thano na reductio no fth ebar k inoculum.Th e fungicideOrth oDifolata nprove dt ob emor e effective thancoppe rbase dcompounds . Culturalpractices ,lik eth eapplicatio no fhig h levelso ffertilizers ,manur e andmulc han drigi dprunin gpractices ,ha dn oeffec to nth eleve lo f C. coffeanum inbranches .Coppe r containing fungicidespushe dth e Colletotrichum balancei n favouro f C. coffeanum. Berries fromnon-coppe r sprayed coffee fieldswer e less susceptible tostandar d conidial suspensions of C. coffeanum thanberrie s from copper sprayed trees.A simila r effecto ffungicide s shouldb econsidere di n Southan dCentra lAmerica n coffeegrowin g countries,wher eth eapplicatio no f fungicidesha sincrease d tremendously sinceth eoccurrenc eo f Hemileia vastatrix Berk,e tBr .i nBrazil . (13tables ;1 2figures ;5 reprinte d papers;22 3references) .

103 12. BIBLIOGRAPHY

Adero,W.E. , 1969.Coffe e Industry of Taita (Cooperative sector).Keny a Coffee 34:42 . Aduayi,E.A. , 1972.Effec t of copper sprays onminera l nutrient content and growth of arabica coffee seedlings inKenya . Soil Sc. PI. Analysis 3:323-328 . Anon., 1964.A handbook onArabic a coffee inTanganyika . (Ed.J.B.D . Robinson), Tan. Coffee Brd., 182pp . Anon., 1970.A nAtla s of Coffee Pests and Diseases.Coffe e Board ofKenya , Nairobi, 122pp . Anon., 1971.Guid e to the Coffee Research Station,Ruiru . (Ed.C . Bould). Coffee Research Foundation,2 8pp . Anon., 1973.Agricultura l Sector Survey- Kenya . Inter.Developm .Ass. ,Vol .I -I I Anon., 1975.Bette r coffee farming. Control of coffee berry disease (CBD)an d leaf rust. Kenya Coffee 40:211-214 . Anon., 1976.Curren t economic position and prospects of Kenya,Repor t No.1284a-KE , World Bank. Arx,J.A .von , 1957.Di eArte n der Gattung Colletotriahwn Cda.Phytopath .Z .29 : 413-468. Arx,J.A .von , 1970.A revision of the fungi classified as Gloeosporium. Verl. J. Cramer,Lehre ,20 3pp . Baker,C.J. , 1972.Fiel d trials with fungicides for CBD control 1970-71.Keny a Coffee 37,p .305 ,307 ,309 . Baker,C.J. , 1973. 1972 trials with new and recommended fungicides.Keny a Coffee 38: 185-191. Bock,K.R. , 1956.Investigation s of coffee berry disease - laboratory studies. E. Afr.Agric .For .J . 22:97-103 . Bock,K.R. , 1959.Note s on theus e of captan sprays in coffee.Keny a Coffee24 : 405-406. Bock,K.R. , 1962a. Seasonal periodicity of coffee leaf rust and factors affecting the severity of outbreaks inKeny a Colony.Trans .Br .mycol . Soc. 45:289-300 . Bock,K.R. , 1962b.Contro l of coffee leaf rust inKeny a Colony.Trans .Br .mycol . Soc. 45:301-313 . Bock,K.R. , 1963.Th e control of coffee berry disease inKenya .Emp .J .Exp . Agric. 31:97-107 . Bock,K.R. , 1970.Plan t pathology inEas t Africa.Rev . Plant Path.49 : 1-6. Bock,K.R . &Rayner ,R.W. , 1956.Contro l of Coffee Berry Disease inKenya . Nature 178:217-218 . Bock,K.R. , Robinson,J.B.D . &Chamberlain ,G.T. , 1958.Zin c deficiency induced by mercury in Coffea arabiaa. Nature 182: 1607-1608. Boerema,G.H . &Bollen ,G.J. , 1975.Conidiogenesi s and conidial septation as differentiating criteria between Phoma and Ascochyta. Personia 8, 2:111-144 . Boisson, C, 1960.L'anthracnos ed u Cafeier.Rev .myc . 25:263-292 . Brown,L.H . & Cocheme,J. , 1969.A study of theagroclimatolog y of the highlands of EasternAfrica ,F.A.O. , Rome,33 0pp . Butler,E.J. , 1918.Fung i and diseases inplants .Calcutta ,Thacker ,Spin k &Co. , 547pp . Butt,D.J . &Butters ,B. , 1966.Th e control of coffee berry disease inUganda . Proc. First Spec.Meetin g of Coffee Res.i nE.Afr. ,Nairobi ,Kenya ,E .Afr . Comm. Serv.Org . Carvalho,A . &Monaco ,L.C. , 1969.Th e breeding of arabica coffee,p . 198-216. In:Outline s of perennial crop breeding in the tropics. (EdsF.P . Ferwerda and F. Wit).Misc .paper s 4,LandbHogesch. ,Wageningen , theNetherlands ,51 1 pp. Chevalier,A. , 1947.Le s cafeiers du globe.Vol . I-III .Encl . Biol.,Paris , P. Lechevalier.

104 Clarke,R.T .& Williamson ,J.G. , 1963.Th e control of coffeeberr y disease in Kenya by seasonal sprays.Emp .J . Exp.Agric . 31:327-333 . Collins,J.A. , 1968.196 7Fiel d trials on control of coffee berry disease in Kenya. 2.Trial s conducted by the Shell Chemical Co. (E.Afr. )Ltd .Keny a Coffee, 33:227-231 . Cook,R.T.A. , 1975.Th e future ofbenzimidazol e fungicides for CBD control in Kenya.Keny a Coffee,40 :139-144 . Cook,R.T.A .& Pereira ,J.L. , 1977.Strain s of Colletotriehum aoffeanum, the causal agent of coffee berry disease,toleran t tobenzimidazol e compounds in Kenya.Keny a Coffee 42,491 :63-76 . Critchett,C , 1969.Som e evidence for apossibl e latent phase of CBD in the green coffee berry. The behaviour of C. aoffeanum on some other tropical fruits.Proc . Fourth Spec.Meetin g on Coffee Res.i nE .Afr .Kampala ,Uganda ,E .Afr .Comm . Serv.Org . Crowe,T.J. , 1959.Th eYellow-Heade d Borer.Keny a Coffee,24 . Crowe,T.J. , 1960.Th e Leaf Skeletonizer.Keny a Coffee,25 :256-257 . Crowe,T.J. , 1962.Th e Star Scale.Keny a Coffee,27 :9-11 . Crowe,T.J. , 1974.Coffe e LeafMiner s inKenya .Keny a Coffee 29,I -III . Crowe,T.J . &Leeuwangh ,J. , 1965.Th e Green Looper.Keny a Coffee,30 . FA0-UNESC0, 1974.Soi lma p of theWorl d 1: 5.000.000 ;Vol . I.,Unesco ,Paris , 60pp . Fernandez,B. , 1961.Diebac k of coffee shoots caused by Phoma and Colletotriehum spp.Cenicaf e 12:127-140 . Fernie,L.M. , 1966.Impression s of coffee inEthiopia .Keny a Coffee,31 :115-121 . Fernie,L.M . &Vermeulen ,H. , 1966.Th e screening of Lyamungamateria l for resistance to coffee berry disease. Proc.Firs t Spec.Meetin g on Coffee Res.i n E. Afr.,Nairobi ,Kenya ,E .Afr .Comm . Serv.Org. : 65-72. Firman, I.D., 1964.Screenin g of coffee for resistance tocoffe e berrydisease . E.Afr .Agric .J . 29:192-194 . Firman, I.D., 1965.A review of leaf rust and coffee berry disease control inKenya . Trop.Agr .Trinida d 42,2 : 111-119. Firman, I.D., 1970.Possibl e side effects of fungicides onbanan a and coffee diseases.Nature ,London ,225 :1161 . Firman,I.D . &Waller ,J.M. , 1977.Coffe e Berry Disease and other Colletotriehum diseases of coffee.C.A.B . Phytopath.Paper sNo . 20,5 3pp . Fordyce,D.M . & Shaw,M.W., 1968. 1967Fiel d trials on control of coffee berry disease inKenya .3 .Trial s conducted byMurph y Chemicals. (E.Afr. )Ltd .Keny a Coffee 33:231-232 . Foucart,G . &Brion ,L. , 1963.L a lutte contre les ennemiesd u cafeier d'Arabie au Rwanda. Bull.Inf . Inst.Nat .Etud eAgron .Cong o Beige 12:141-152 . Furtado, I., 1969.Effec t of copper fungicides on the occurrence of the pathogenic form of Colletotriehum aoffeanum. Trans Br.mycol . Soc. 53:325-328 . Furtado, I., 1970.Th e effect of copper fungicides upon the barkmicroflor a of Coffea avabiaa with particular reference to thepathogeni c strainso f Colletotriehum aoffeanum. Ph.D .Thesis ,Univ .E .Afr . Gassert,W.L. , 1976.Zu r Epidemiologie derKaffeekirschen-krankhei t ('Colletotriehum aoffeanum Noack sensu Hindorf) inXthiopien .Ph .D . Thesis,Justus-Liebi g Univ. Giessen, 120pp . Gassert,W.L. , 1978.Rind eun d mummifizierte Kirschen alsQuell e furda s Primarinokulumde sKaffeekirschen-krankhei t inXthiopien . Z.Pflkrankh . Pflschutz 85 (1):30-40 . Gibbs,J.N. , 1969.Inoculu m sources for coffee berrydisease .Ann.appl .Biol .64 : 515-522. Gibbs,J.N. , 1971.Som e factors influencing theperformanc e of spray programmes for thecontro l of coffee berry disease.Ann . appl. Biol.67 :343-356 . Gibbs, J.N., 1972.Effect s of fungicides on the population of Colletotriehum and other fungi inbar k of coffee.Ann .appl .Biol .70 :35-47 .

105 Griffiths,E. , 1970.Contro l of coffee berry disease and leaf rust in 1970. Kenya Coffee,35 :45-47 . Griffiths,E. , 1971. 'Negative'effect s of fungicides incoffee .Trop . Sc.14 : 79-89. Griffiths,E .& Furtado ,I. , 1972.A berr y infection technique for assessment of theCB D straino f Colletotrichum ooffeanwn on coffee branches.Trans .Br .mycol . Soc. 58:313-320 . Griffiths,E .& Gibbs ,J.N. , 1969.Earl y seasons sprays for thecontro l of coffee berry disease.Ann .appl . Biol.64 :523-532 . Griffiths,E. ,Gibbs ,J.N . &Waller ,J.M. ,1971 .Contro lo f coffee berrydisease . Ann. appl.Biol .67 :45-74 . Griffiths,E .& Waller ,J.M. , 1971.Rainfal l and cropping patterns inrelatio n to coffee berry disease.Ann .appl .Biol .67 :75-91 . Haarer,A.E. , 1962.Coffe e Growing.Oxfor d Univ.Press ,London-Ne wYork-Toronto : 127pp . Henrard,P. , 1957. Colletotrichum ooffeanwn Noack.Agric .Louvain , Ser.2 ,5 :39-55 . Heyer,J . &Waweru ,J.K. , 1976.Th eDevelopmen t of the Small Farms Areas,187-221 . In:Agricultura l Development inKenya .A nEconomi c Assessment (EdsHeyer ,Maith a and Senga). Oxford Univ. Press,Nairobi ,Kenya :37 2pp . Hill,M.F. , 1956.Planters 'Progress ,th e Story of Coffee inKenya ,Nairobi . Hindorf,H. , 1970. Colletotrichum spp.isolate d from Coffea arabica L. inKenya . Z.PflKrankh.Pflschut z 77:328-331 . Hindorf,H. , 1972.Qualitativ e und quantitative Unterschiede inde r Colletotrichum- Population auf Coffea arabica L. inKenya .Ph .D . Thesis,Justus-Liebi g Univ. Giessen,Germany , 146pp . Hindorf,H. , 1973a. Colletotrichum population on Coffea arabica inKenya .I .A method for the systemic separation of fungus populations.Phytopath . Z.77 :97-11 6 Hindorf,H. , 1973b. Colletotrichum population on Coffea arabica inKenya . II.Qual ­ itative and quantitative differences in the Colletotrichum population.Phytopath . Z. 77:216-234 . Hindorf,H. , 1973c. Colletotrichum population on Coffea arabica inKenya . III.Th e distribution of Colletotrichum species ondifferen t parts of thecoffe ebush . Phytopath.Z . 77:324-338 . Hindorf,H. , 1973d.Nebenwirkunge n vonKupferhaltige nFungicide n imKaffeeba u von Kenia.Meded . Rijksfac.Landbwsch .Univ . Gent 38:853-856 . Hindorf,H. , 1974. Colletotrichum species from coffee growing areas of theKiamb u district ofKenya .Z .Pflkrankh.Pflschut z 81:108-113 . Hindorf,H. , 1975. Colletotrichum occurring on Coffea arabica: areview .J . Coffee Res. 5 (3/73):43-56 . Hindorf,H . &Muthappa ,B.N. , 1974.A comparison of Colletotrichum coffeanum Noack from South India and Kenya.Phytopath .Z . 80:9-12 . Hocking,D. , 1966.Brow nbligh t {Colletotrichum coffeanum Noack)o fArabic a coffee inEas tAfrica .Ann .appl .Biol .58 :409-421 . Hocking,D. , 1967.Fungicide s for arabica coffee.Preliminar y relationship among somene w fungicides,lea f rust -Hemileia vastatrix - and leaf fall.Trop .Agric . 44: 83-88. Hocking,D. , 1971.Alternativ e hosts for two races of Colletotrichum coffeanum from coffee.Turrialb a 21:234-235 . Hocking,D. , Johanns,J.C .& Vermeulen ,H. , 1967.Ascospor e production,discharg e and infection by Glomevella cingulata causing coffee berry disease.Natur e214 , 5093: 1144-1145. Hollies,M.A. , 1967.Chroni c leaf fall inarabic a coffee inTanzania .E .Afr . Agric.For .J . 32:404-410 . Huxley,P.A. ,Turk ,A .& Mitchell ,H.W. , 1969.Confidenc e Limits of Expected Rainfall inCoffe e Areas inKenya .Par t I.Keny a Coffee 34:57-61 . Jones,P.A. , 1956.Note s onvarietie s of Coffea arabica inKenya , 158-166.In : Selected Articles onCoffe e Culture,Coff .Brd . ofKenya ,CRS ,Ruiru .

106 Jones,P.A. ,Robinson ,J.B.D . &Wallis ,J.A.N. , 1961.Fertilizers ,manur e and mulch inKeny a coffee growing.Keny a Coffee 26:441-459 . Kieran,J.A. , 1969.Th e origin of commercial arabica coffee production inEas t Africa.Afr .Historica l Studies 21:51-67 . Krug,C.A . &D e Poerck,R.A. , 1968.Worl d Coffee Survey. FAOAgric . StudiesNo . 76, 476pp . Lumb,F.E. , 1966.Variation s of rainfall over LakeVictori a Catchment since 1899 and over East Africa since 1934.Keny a Coffee 31:347-350 . Maina,J.W. , 1969.Agricultur e and land tenure inKenya : 177-197.J_n :Eas tAfrica : itspeopl e and resources (Ed.W.T.W .Morgan) , Oxford Univ. Press,Nairobi ,Kenya , 312pp . Maitha,J.K. , 1976.Th e Kenyan Economy: 33-67.J_n :Agricultura l Development in Kenya. AnEconomi c Assessment (EdsHeyer ,Maith a and Senga), Oxford Univ.Press , Nairobi,Kenya ,37 2pp . McDonald,J. , 1921.Fungoi d diseases of coffee inKeny a Colony,Col .7 ,Prot .o f Kenya, Bull.Dept .o fAgric . McDonald,J. , 1922.Coffe e Berry Disease.Mimeographe d circular. McDonald,J. , 1925.Ann .Rep .Dept .o fAgric ,Keny a 1924:142 . McDonald,J. , 1926.A preliminar y account of adiseas e ofgree n coffee berries in Kenya Colony. Trans.Br .mycol . Soc. 11:145-154 . McDonald,J. , 1930.Ann .Rep .Dept .o f Agric,Keny a 1929: 210-224. McDonald,J. , 1931.Ann .Rep .Dept .o fAgric ,Keny a 1930: 119-120. McDonald,J. , 1932.Ann .Rep .Dept .o fAgric ,Keny a 1931: 119. McDonald,J. , 1936.Th e susceptibility of Harer coffee todiseases .Mon .Bull . Coffee Board Kenya, II,22 : 191 (R.A.M. 14:171) . Meiffren,M. , 1957.Le smaladie s du Cafeier enCot e d'Ivoire,Abidjan ,Haut-Comm . de l'A.O.F. Meulen,A .va nder , 1939.Ove r de bouw end e periodieke ontwikkeling derbloemknop - penb yCoj\fea-soorten .Verh .Kon .Ned .Akad .Wetensch . 2e Sectie,38-2 :1-128 . Moens,P. , 1962.Etud e ecologique du developpement generatif et vegetatif des bourgeous de Coffea oanephora Pierre.Publ. INEAC,Ser . sci.96 , 103pp . Moens, P., 1963.Le sbourgeou svegetati f etgeneratif sd e Coffea oanephora Pierre. La Cellule (Louvain)63 :165-244 . Mogk,M ., 1970 .Th e development of the coffee cherry and its susceptible stages to berry disease. 18th Int.Hort .Congr. ,Te lAviv ,Israel . Mogk,M. , 1973.Untersuchunge n zurEpidemiologi evo n Colletotriahum aoffeanwn Noack sensu Hindorf inKenia ,ein e Analyse derWirt-Parasit-Umwelt-Beziehungen . Thes. Justus-Liebig Univ.,Giessen , 163pp . Mogk,M . &Hindorf ,H. , 1975.Verlust e durchdi eKaffeekirschenkrankhei t (Colletotriahum ooffeanum) inverschiedene n Stadien der Kirschen-Entwicklung. Z.Pflkrankh.Pflschut z 82:193-200 . Mulder,D . &Hocking ,D. , 1967.Hypothesi s to explain theuneve n distribution of coffee berry disease inarea s of endemic occurrence.Meded .Rijksfa c Landbwsch. Univ.Gen t 32:729-73 1. Mulinge, S.K., 1970a.Developmen t of coffee berry disease inrelatio n to the stage of growth of theberry .Ann .appl . Biol.65 :269-276 . Mulinge, S.K., 1970b.Variation s in the level of Colletotriahum aoffeanwn Noack in thebar k of Coffea arabica cultivars.E .Afr .Agric .For .J . 36:227-230 . Mulinge, S.K., 1971a.Distributio n of Colletotriahum eoffeanum strainswithi n coffee trees.Trans .Br .mycol .Soc .56 :478-480 . Mulinge, S.K., 1971b.Effec t of altitude on thedistributio n of the fungus causing coffee berry disease inKenya .Ann .appl .Biol .67 :93-98 . Mulinge, S.K., 1973. Outbreaks and new records inEthiopia ,coffe e berrydisease . FAO Plant Protection Bulletin 21:85-86 . Mulinge, S.K. &Griffiths ,E. , 1974a.Effect s of fungicides on leaf rust,berr y disease, foliation and yield of coffee.Trans .Br .mycol .Soc .62 :495-507 .

107 Mulinge, S.K. &Griffiths ,E. , 1974b.Effect s of temperature on the Colletotriahum population oncoffe e bark.Trans .Br .mycol . Soc. 62:610-614 . Muller,R.A. , 1964.L'anthracnos e desbaie s du cafeier d'arabie (Coffea arabica) duea Colletotriahum coffeanum Noack au Cameroun. I.F.C.C.Bull .no .6 : 9-38. Muller,R.A. , 1968.L a lutte contre l'anthracnose desbaie s du cafeier Arabica due aun e formed e Colletotriahum coffeanum Noack auKenya .Cafe-Cacao-Th eXII , 1:39-52 . Muller,R.A. , 1970.L'evaluatio nd e l'anthracnose desbaie s du cafeier d'arabie (Coffea arabica) duea un eform ed e Colletotriahum coffeanum Noacka uCameroun . Cafe-Cacao-The 14:114-129 . Muller,R.A. , 1973.L'anthracnos e dubaie s du cafeier d'arabie (Coffea arabica) due aun e formevirulent e du Colletotriahum coffeanum Noack. I.Variation s de la sensibilite des fruits au coursd e leurdevelopment . II.L'irrigatio n methode preventive de controle de lamaladie . Cafe-Cacao-The 17:281-312 . Muller,R.A. , 1978.Contributio n a la Connaissance de laPhytomycocenos e Coffea arabica L., Colletotriahum coffeanum Noack sensu Hindorf, Hemileia vastatrix Berk, etBr. , Hemileia ooffeicola Maubl. etRoger .Thes .Univ .Pierr e &Mari e Curie,Paris ,30 4pp . Muller,R.A . &Gestin ,A.J. , 1967.Contributio n a lamis e aupoin t demethode s de lutte contre l'anthracnose debaie s du cafeier arabie (Coffea arabica) duea une formed e Colletotriahum coffeanum au Cameroun.Cafe-Cacao-Th e 11:157-178 . Muthappa,B.N. , 1970.Studie s on the role of Colletotriahum coffeanum incausin g stalk rot of leaves and berries of Arabica coffee.India n Coffee 34:263-264 . Neely,D. , 1970.Resistanc e of foliarprotectan t fungicides.Phytopatholog y60 : 1583-1596. Neely,D. , 1971.Depositio n and tenacity of foliage protectant fungicides.PI .Dis . Reptr 55:898-901 . Nicholls,W. , 1969.Th e progress of bark formation inarabic a coffee.Keny a Coffee 34:429-434 . Noack,F. , 1901.Di eKrankheite n des Kaffeebaumes inBrasilien . Z.Pflkrankh .11 : 202. Nutman,F.J. , 1966.Coffe eberr y disease -a review.Proc .Firs t Spec.Meetin g on CoffeeRes .i nE .Afr, ,Nairobi ,Kenya ,E .Afr .Comm . Serv.Org . Nutman,F.J . &Roberts ,F.M. , 1960a.Investigation s on adiseas e of Coffea arabica caused by a form of Colletotriahum coffeanum Noack. I. Some factors affecting infection by thepathogen .Trans .Br .mycol .Soc .43 :489-505 . Nutman,F.J . &Roberts ,F.M. , 1960b.Investigation s on adiseas e of Coffea arabica caused bya form of Colletotriahum coffeanum Noack. II.Som e factors affecting germination and infection and their relationship todiseas e distribution.Trans . Br.mycol .Soc .43 :643-659 . Nutman,F.J . &Roberts ,F.M. , 1961.Investigation s on adiseas e of Coffea arabica caused by afor m of Colletotriahum coffeanum Noack. III.Th e relationship between infection of bearing wood and disease incidence.Trans .Br .mycol .Soc .44 : 511-521. Nutman,F.J . &Roberts ,F.M. , 1962a.Coffe e berry disease and leaf rust research. Kenya Coffee 27: 13-17; 24-25; 110-111;236-240 ;273 . Nutman,F.J . &Roberts ,F.M. , 1962b.Stimulatio n of twopathogeni c fungi by high dilutions of fungicides,Trans .Br .mycol .Soc .45 :449-456 . Nutman,F.J . &Roberts ,F.M. , 1969a.Seasona lvariation s in the sporulating capa­ city of the fungus causing coffee berry disease.Ann .appl . Biol. 64:85-99 . Nutman,F.J . &Roberts ,F.M. , 1969b.Th e effect of fungicidal treatments on sporulating capacity inrelatio n to the control of coffeeberr ydisease .Ann . appl. Biol. 64:101-112 . Nutman,F.J . &Roberts ,F.M. , 1969c.Th e stimulating effect of some fungicides on Glomerella cingulata inrelatio n to the control of coffee berry disease.Ann . appl. Biol. 64:335-344 .

108 Nutman,F.J . &Roberts ,F.M. , 1969d. Climatic conditions inrelatio n to the spread of coffeeberr y disease since 1962 in theEas tRif t districts ofKenya .E .Afr . Agric.For .J . 35: 118-127. Nutman, F.J. &Roberts ,F.M. , 1969e.Folia r nutrients inrelatio n to coffee berry disease.E .Afr .Agric .For .J . 35:217-223 . Nutman,F.J . &Roberts ,F.M. , 1970.A note on a laboratory technique for testing fungicides against coffeeberr ydisease .E .Afr .Agric .For .J . 35:225-228 . Nutman,F.J. , Roberts,F.M . &Vermeulen ,H. , 1968.Coffe e berry disease.Ann . Rep. 1967.E .Afr .Agric .For .Res .Org. ,Recor d of Research: 96-98. Okioga,D.M. , 1976.Occurrenc e of strains of Colletotriohvm ooffeanvm resistant to methyl benzimidazol-2-ylcarbamate (carbendazim)an d chemically similar compounds. Ann. appl, Biol.84 :21-30 . Okioga,D.M . &Mulinge ,S.K. , 1974. 1973 trialswit h the newan d recommended fungicides.Keny a Coffee 39:319-322 . Patel,R.Z. , 1971.Redistributio n of captafol.Ann .Rep .Coff .Res .Stat. ,Kenya , 1970-1971,5 6pp . Pereira,H.C . &Jones ,P.A. ,1954 .Fiel d responses byKeny a coffee tofertilizers , manures andmulches .Emp . J. Exp.Agric .2 2 (85):23-36 . Pereira,J.L. ,Mapother ,H.R. , Cooke,B.K . &Griffiths ,E. , 1973.Redistributio n of fungicides incoffe e trees.Exp .Agric .9 :209-218 . Ponte,A.M . da, 1966.Sprayin g Arabica coffeewit h calcium super phosphates for the control of coffee berry disease,normall y attributed to Colletotriohwn ooffeanvm Noack. F.A.O. Conf.,Ri o deJan. ,Brazi l (reprinted inKeny a Coffee 31:21-22) . Powell,H. , 1908/1908.Coffe e (Coffea arabica var.). Agric.J . Br.E .Afr . 1: 132-142. Purseglove,J.W . (Ed.), 1968. Coffea L._In :Tropica l Crops,Dicotyledons,Vol .2 , 458-492.Longmans.Ltd. ,London ,Gree n &Co . Ramos,A.H .& Shavdia,L.D. , 1976.A dieback of coffee inKenya . PI. Dis.Rept r 60:831-835 . Rayner,R.W. , 1941.Repor t of Plant Pathologist (Coffee Service). Report of the Department ofAgriculture ,Kenya .Unpublished . Rayner,R.W. , 1948.Laten t infection of Coffea arabica. Nature 161:245-246 . Rayner,R.W. , 1952.Coffe e berry disease.A survey of investigations carried out up to 1950.E .Afr .Agric .J . 17:130-158 . Rayner,R.W. , 1957.Toni c copper spraying of coffee.Proc .Nairob i Scient.Phil . Soc. 9: 12-16. Rayner,R.W. , 1960.Rus t disease of Coffee.Wrld . Crop 12,1 6pp . Rayner,R.W. , 1962.Th e control of coffee rust inKeny a by fungicides.Ann .appl . Biol. 50:245-261 . Rayner,R.W . &Jones ,P.A. , 1948.Toni c copper spraying.Mon .Bull .Coffe e Bd. Kenya,Februar y issue. Richardson,L.T .& Thorn ,G.D. , 1962.Stimulatio n of Spore Germination and Growth of Glomerella oingulata by Copper and otherHeav yMeta l Ions.Phytopatholog y 52: 865-869. Robinson,R.A. , 1976.Plan t Pathosystems. Springer-Verlag,Berlin ,Heidelberg , NewYork .Adv . ser.Agric . Sc. 3, 184pp . Robinson,R.A . &Waller ,J.M. , 1966.A phyllosphere hypothesis concerning coffee. Proc.Firs t Spec.Meetin g on Coff.Res .i nE .Afr .Nairobi ,Kwnya ,E .Afr .Comm . Serv.Org . Saccas,A.M . &Charpentier ,J. , 1969a.L'anthracnos e des Cafeiers robustae t excelsa due a ColZetotrichum ooffeanvm Noack enRepubliqu e Centrafricaine. Cafe-Cacao-The 13:131-150 . Saccas,A.M . & Charpentier,J. , 1969b.L'anthracnos e des".Cafeiersrobust a et excelsa due a Colletotvichvm ooffeanvm Noack enRepubliqu e Centrafricaine. Cafe-Cacao-The 13:221-230 . Senga,W.M. , 1976.Kenya' sAgricultura l Sector:69-110 ._In :Agricultura l Develop­ ment inKenya .A nEconomi c Assessment (EdsHeyer ,Maith a and Senga). Oxford Univ. Press,Nairobi ,Kenya ,37 2pp . 109 Shaw,D.E. , 1967.Diffus e yellow leaf spot ofArabic a coffee inPapu a and New Guinea. Papua New.Agric .J . 18:120-121 . Sheffield,F.M.L. , 1963.Stem-pittin g in Coffea arabica. Ann.appl . Biol.52 : 211-216. Siddiqi,M.A. , 1965.A not e on coffee bark disease inEas tAfrica .Keny a Coffee30 . Simmonds,J.H. , 1963. Studies on the latent phase of Colletotriahum causing ripe rots of tropical fruits.Queensl .J . Agric.& Anim . Sc. 20:373-424 . Small,W. , 1926.O n the occurrences of a species of Colletotriahum. Trans.Br . mycol. Soc. 11:112-137 . Smith,L.D. , 1976.A n overview ofAgricultura l Development Policy, 111-151.In : Agricultural Development inKenya .A n Economic Assessment. (EdsHeyer ,Maith a and Senga). Oxford Univ. Press,Nairobi ,Kenya ,37 2pp . Soil Survey Staff, 1975.Soi l taxonomy; Basic system of Soil Classificationfo r making and interpreting soil surveys,U.S.D.A. , Soil Cons.Serv. ,Agric .Handb . 436:Wash .D.C . Sombroek,W.G. , 1979.Th e soilma p of Kenya at scale 1: 500.000 .Misc .Paper , Kenya Soil Survey,Nairobi ,i npress . Steiner,K.G. , 1972a.Th e effect of foliar fertilizers on thedevelopmen t of CBD and on theyiel d of coffee.Keny a Coffee 36:160 . Steiner,K.G. , 1972b.Th e influence of surfacewa x obtained from green berries of six selections of Coffea arabica on the germination of Colletotriahum ooffeanum. Kenya Coffee 37:179 . Steiner,K.G. , 1973a.Zu r Frage der Infektionsperiodenbe i derKaffeekirschen - Krankheit(Colletotriahum ooffeanum Noack), Z. Pflkrankh.Pflschutz 80:492-497 . Steiner,K.G. , 1973b.Versuch e zurErmittlun g derWirkungsdaue r vonFungicide n bei Colletotriahum ooffeanum Noack,de mErrege r der Kaffeekirschen-Krankheit.Z . Pflkrankh.Pflschutz 80:532-546 . Steiner,K.G. , 1973c.De r Einfluss der Fungizidbehandlung auf den Verlauf der Kaffeekirschen-Krankheit.Z . Pflkrankh.Pflschutz 80:671-681 . Steiner,K.G. , 1973d.'Ertragssteigerungen imKaffeeba u durch diekombiniert e Anwendung vonFungicide n und Blatt-dungern.Z , Pfl.Ern.Bodenk . 137:81-85 . Steiner,K.G. , 1974.Di e Sporulation von Colletotriahum ooffeanum Noack auf Kirschenvo n Coffea arabica L. Phytopath.Z . 79: 179-189. Stephan,B.R. , 1968.Untersuchunge n zumNachwei s der Hetero-karyose bei Colletot­ riahum gloeosporioides Penzig unter Verwendung auxotropher Mutanten.Zentr . bl. Baht.Paras .Ingkrankh .Hyg . 122:420-435 . Steward,R.B. , 1957.Lea f blight and stemdiebac k of coffee caused by anun - described species ofAscochyta .Mycologi a 49:430-433 . Strobal,J.W. , 1960.Physiologicall y induced morphologic and pathogenic variations of some Glomerella cingulata isolates.Diss .Abstr . 21:1014-1015 . Swynnerton,R.J.M. , 1954.A Plan to Intensify theDevelopmen t ofAfrica nAgricul ­ ture inKenya .Governmen t Printer,Nairobi . Tapley, R.G., 1960.Th ewhit e coffee borer Anthores leuconotus Pasc. and itscon ­ trol. Bull.Ent .Res .5 1 (2):279-301 . Tapley,R.G. , 1964.Coffe e berry disease inTanganyika . Tan.Coffe eNew s 38:40-45 . Thorold, C.A., 1945.Elgo ndiebac k disease of coffee.E .Afr .Agric .J . 10:198-206 . Trench,A.D. , 1926.Som eAspect s of Coffee Production inKenya .Proc . E.Afr .Conf . onAgric. :49-55 . Tuite,J. , 1969.Plan t pathological methods.Fung i and Bacteria.Burg .Publ .Co. , U.S.A., 239pp . Twist,T.K . &Coltman ,W.F. , 1966.Organi cmatte r and healthy coffeewit hrefer ­ ence to coffee berry disease and quality.Keny a Coffee 31:261-263 . Vargas, E.& Gonzales ,L.C. , 1972.L amanch amantecos a des cafe causada por Colletotriahum spp.Turrialb a 22:129-135 . Vermeulen,H. , 1965a.Wor k inprogres s inth e Coffee Research Service.Keny aCoffee , 30, Febr.,2 pp . Vermeulen,H. , 1965b.Coffe e Diseases inKenya .Firs t session of the F.A.O.

110 technicalworkin g party on coffee production and protection,Ri o deJaneiro , Brazil.3r d Techn.Session ,n °Ce/65/14 . Vermeulen,H. , 1966a.Applicatio n of Superphosphates to Control Coffee Berry Disease inKenya .Keny a Coffee 31:Febr. , 1p . Vermeulen,H. , 1966b.Screenin g of fungicides for the control of coffeeberr y dis­ ease. Proc.Firs t Spec.Meetin g on Coffee Res.i nE .Afr .Nairobi ,Kenya ,E .Afr . Comm. Serv.Org. :22-3 7(mimeographed) . Vermeulen,H. , 1968.Screenin g of fungicides for control of coffee berry disease inKenya .Exp .Agric .4 :255-261 . Vermeulen,H. , 1970a.Coffe e berry disease inKenya . I. Cotletotrichwn spp. colonizing thebar k of Coffea arabica. Neth.J . Plant Path. 76:277-284 . Vermeulen,H. , 1970b.Coffe e berrydiseas e inKenya .II .Th e role of Glomerella cingulata in the ColZetotriahum population colonizing thebar k of Coffea arabica. Neth.J . Plant Path.76 :285-292 . Vermeulen, 1970c.Coffe eberr ydisease .Ann .Rep . 1969.E .Afr .Agric .For .Res . Org.,Recor d of Research:92-93 . Vermeulen,H . &Mehlich, A. , 1966.Effect s onp Han d Soluble Phosphorus of Adding SomeAgricultura l Chemicals to Superphosphate Spray Solution.Keny a Coffee31_ , 2pp . Vermeulen,H . &Patwa ,N.T. , 1966.A shortnot e on some spray timing trials against coffee berrydisease .Keny a Coffee 31.,3 pp . Vine, B.H.& Griffiths ,E. , 1968.A bio-assaymetho d for determining theper ­ sistence of captafol residues oncoffe e leaves.E .Afr .Agric .& For .J .34 : 160-163. Vine,B.H. ,Vine ,P.A .& Griffiths ,E. , 1973a.Evaluatio n of fungicides for the control of coffee berry disease inKenya .Ann .appl . Biol.75 :359-375 . Vine, B.H.,Vine ,P.A .& Griffiths ,E. , 1973b.Som e problems of evaluating fun­ gicides for use oncoffe e inKenya .Ann .appl .Biol .75 :37-385 . Vossen,H.A.M .va n der, 1973.Coffe e breeding inKenya .Keny a Coffee 38:253-256 . Vossen,H.A.M .va nder , 1974.Plan t breeding.Ann .Rep .Coffe eRes .& Found. , Kenya, 1973/1974:40-51 . Vossen,H.A.M .va nde r & Cook,R.T.A. , 1975.Incidenc e and control ofberr y blotch caused by Ceroospora coffeicola on arabica coffee inKenya .Keny a Coffee40 : 58-61. Vossen,H.A.M .va n der,Cook ,R.T.A . &Murakaru ,G.N.W. , 1976.Breedin g forre ­ sistance tocoffe e berry disease caused by Colletotvichvm coffeanum Noack (sensu Hindorf) in Coffea arabica L. I.Method s ofpre-selectio nfo rresistance . Euphytica 25:733-745 . Waller,J.M. , 1971.Th e incidence of climatic conditions favourable tocoffe e berry disease inKenya .Exp .Agric .7 :303-314 . Waller,J.M. , 1972a.Water-born e spore dispersal incoffe e berrydiseas e and its relation tocontrol .Ann .appl .Biol .71 :1-18 . Waller,J.M. , 1972b.Coffe erus t inLati nAmerica .PAN S 18,4 :402-408 . Wallis,J.A.N. , 1964.Coffe eberr ydiseas e - 1964.Keny a Coffee29 :401-405 . Wallis,J.A.N . &Firman , I.D., 1965.Sprayin g arabica coffee for the control of coffee berrydisease .Ann .appl .Biol .55 :139-148 . Wallis,J.A.N .& Firman ,I.D. , 1967.A comparison of fungicide sprayvolume s for the control of coffeeberr y disease.Ann .appl .Biol .59 :111-122 . Wallis,J.A.N .& Wormer ,Th .M ., 1971 .Kaffee ,496-511 . In:Handbuc hde r Land- wirtschaft und Ernahrung inde nEntwicklungslandern . (EdsVo n Blanckenburg und Cremer). Band 2,Verl . EugenUlmer ,Stuttgart . Wellman,F.L. , 1954.Fung i isolated most frequently fromdyin g coffee fruit branches.Phytopatholog y 44:509 . Wellman,F.L. , 1961.Coffee ;botany , cultivation and utilization.London ,Leonar d Hill, 488pp . Wellman,F.L. , 1970.Th e rust, Hemileia vastatrix, now firmly established on coffee inBrazil .PI . Dis.Rept r 54:539-541 .

111 Wheatley,P.E. , 1962.Antesti a Testing.Keny a Coffee 27:405 . Wheatley,P.E. , 196A.Th eGran t Looper.Keny a Coffee 29,5 pp . Wheeler,H.E. , 1954.Genetic s and evolution ofheterothalis m in Glomerella. Phytopathology 44:342-345 . Woodhead,T. , 1968.Micro-meteorologica l studies of coffee berry disease;a fiel d investigation into the incidence of thephysica l conditions favourable for spore germination.Ann . appl. Biol. 62:451-463 . Wormer,Th .M. , 1964.Th e growth of the coffee berry.Ann .o f Bot.28 :47-55 . Wormer,Th .M . &Gituanja ,J. , 1970.Flora l initiation and flowering of Coffea arabiaa L. inKenya .Exp .Agric . 6: 157-170. Wormer,Th .M .&Njuguna , S.G., 1966.Bea n size and shape asqualit y factors in Kenya coffee.Keny a Coffee31 .

Niets isvolmaakt ,niet s iste neinde , elkeverstarrin g isbedrog .

Fragment wit: Apolloon tegenover XVII, Herwig Hensen.

112 CURRICULUM VITAE

HansVermeule nwer do p1 4maar t 1932t eBandung ,Indonesi egeboren .He tbelang - rijkste deelva nzij nmiddelbar eschoolopleidin g vondplaat so phe tBaarn s Lyceum, alwaarhi ji n195 2he teinddiplom aH.B.S.- Bbehaalde .Daarn avolgd ed estudi eaa n deLandbouwhogeschoo lt eWageningen ,di ehi jo p7 apri l196 0afsloo tme the tbe - halenva nhe tingenieursdiplom ai nd erichtin g tuinbouwplantenteeltme tal shoofd - vakd etuinbouwplantenteel t enal sbijvakke nd efytopathologie ,d evirologi ee nd e tropische plantensystematiek. Van196 0to tein d196 2wer di nSurinam eee nonderzoe kverrich tnaa rd ever - wekkerva nd ezeefvatenziekt eva nliberi akoffie ,onde rauspicie nva nd eStichtin g WetenschappelijkOnderzoe k Surinamee nd eNederlands eAntille n (WOSUNA). N aterug - keeri nNederlan dvolgd ebegi n196 3ee naanstellin gbi jd ePlantenziektenkundig e Dienstt eWageningen . Inaugustu s 1964aanvaardd ehi jee nbetrekkin gal sfytopatholoo g aanhe t koffieproefstationt eRuiru ,Kenia .I n196 6wer dbeslote nhe tverblij fi nKeni at e verlengen,i nhe traa mva nd ebilateral e samenwerking tussenKeni ae nNederland . De detacheringbi jd eOostafrikaans e land-e nbosbouwkundig e onderzoekorganisatie (EastAfrica nAgricultur ean dForestr y ResearchOrganization )t eNairob ivolgd ei n januari196 7me tal sopdrach the ti nRuir ubegorme nonderzoe k verdervoor tt e zetten.Di tonderzoe kwer d inSeptembe r 1969afgesloten . In196 9wer dhi jopgenome ni nd epoo lva nlandbouwdeskundige nva nhe t Ministerieva nLandbou we nVisseri j inNederland .N aterugkee rui tKeni avolgd ei n 1970uitzendin g naar Indonesie,opnieu wi nhe traa mva nee nbilateraal ,agrarisc h samenwerkingsproject.Gedurend e ruim zevenjare nwerkt ehi jaldaa ral sadviseu r metal svoorn;»amst eopdrach tee nafdelin g gewasbeschermingo pt ebouwe nbinne nhe t Indonesische instituutvoo r tuinbouwkundigonderzoek .