Entomopathogenicfung i against whiteflies
Tritrophic Interactions betweenAschersonia species, Trialeurodes vaporariorum and Bemisia argentifolii, and glasshouse crops
EllisT.M .Meeke s Promoter: Prof.Dr . J.C.va n Lenteren Hoogleraar in deEntomologi e Wageningen Universiteit
Copromotor: Dr.Ir .J.J .Franse n Senior onderzoeker Entomologie Proefstation voorBloemisteri j enGlasgroente s Aalsmeer
Samenstellingpromotiecommissie : Dr.J.K .Pel l (IACRRothamsted , UK) Dr.W .Gam s (Centraal Bureau Schimmelcultures, NL) Prof.Dr .J.M . Vlak (Wageningen Universiteit) Prof.Dr .Jr .A.H.C . vanBrugge n (Wageningen Universiteit) . 1 I /
E.T.M. Meekes
Entomopathogenic fungi against whiteflies
Tritrophic interactions betweenAschersonia species, Trialeurodes vaporariorum andBemisia argentifolii, and glasshouse crops
Proefschrift terverkrijgin g van degraa dva n doctor opgeza gva nd erecto r magnificus van deWageninge n Universiteit Prof.Dr .Ir .L . Speelman inhe t openbaar te verdedigen opwoensda g 27jun i 2001 desnamiddag st e 13.30i n de Aula
0 D'-? / On _'•v J r\j Meekes,E.T.M E.T.M.(2001) (2001). Entomopathogenicfung i againstwhitefly : tritrophicinteraction sbetwee nAschersonia species , Trialeurodes vaporariorum andBemisia argentifolii, an d glasshouse crops
ISBN: 90-5808-443-4 Contents
Chapter 1 General Introduction 1
Chapter2 Virulence ofAschersonia spp .agains t whiteflies 19 Bemisia argentifoliian d Trialeurodes vaporariorum
Chapter 3 Dosean dtim edependen t mortality of thewhiteflie s 35 Bemisiaargentifolii an d Trialeurodes vaporariorum byAschersonia spp.
Chapter4 Germination andinfectio n behaviour ofAschersonia spp .o n 49 Bemisiaargentifolii an d Trialeurodes vaporariorum on poinsettia
Chapter 5 Persistence of thefunga l whitefly pathogenAschersonia aleyrodis, 69 onthre edifferen t plant species
Chapter 6 Roleo fphyllospher e climate ondifferen t hostplant s inth e 81 interaction between entomopathogenic fungi andth e two pest species Trialeurodes vaporariorum andAphis gossypii
Chapter 7 Relative humidity andhost-plan t speciesinfluenc e 99 mycosiso f whitefly, Trialeurodesvaporariorum, byAschersonia aleyrodis,A. placenta and Verticillium lecanii
Chapter 8 Effect of gerbera cultivar onth e efficacy of entomopathogenic 113
fungi against Trialeurodes vaporariorum
Chapter 9 General Discussion 125
References 145 Summary 163 Samenvatting 167 Nawoord 177 Listo f publications 179 CurriculumVitae 181 1 General introduction
The research discussed in this thesis aimed at evaluating entomopathogenic fungi for biological control of whiteflies. In this chapter I first summarize whitefly pests, damage, biology and its control. Secondly, the background concerning entomopathogenic fungi and whitefly will be described. Finally, the aim of the project and the outline of the thesis is presented.
Whitefly pestsan d damage About 1200 whitefly species have been described, but only few of them are considered as pests.Amon gth elatte rar eth egreenhous ewhitefly , Trialeurodes vaporariorum (Westwood) and, more recently, the silverleaf whitefly, Bemisiaargentifolii (Bellows & Perring) (syn. B. tabaci biotype B).B. argentifoliii spar t of theBemisia tabaci specie s complex, of which manybiotype s are described. The discussion whether B. argentifoliii s a separate species or not is still ongoing (e.g. De Barro& Hart , 2000;Gerling , 2000;Marti n etal., 2000) .I n this thesisth enam e B. argentifoliii s used. Thegreenhous ewhitefl y isa ke ypes to fman ygreenhous evegetable san dornamentals , whereasth e silverleaf whitefly hasbee n aseriou spes ti nfiel d cropsworldwid e since the mid seventies (Byrne etal., 1990a).Bot h species are highlypolyphagous , occur world wide and lead to serious economic losses (Gerling, 1990). The damage to crops is manifold: 1) Direct damage to plant foliage is caused by feeding ofwhiteflies . Adultan dimmatur ewhiteflie s arephloe mfeeders , usingthei rpiercing - sucking mouthparts to introduce digestive juices and to remove chlorophyll and starch. Infested plants may drop leaves prematurely and have reduced plant growth and vigour (Oetting &Buntin , 1996).2 )A by-produc to f feeding isth e excretion ofexces ssugar s in the form of honeydew. It accumulates on the upper surfaces of plant parts where it serves as a substrate for sooty moulds (e.g.Capnodium spp.) , thus reducing leaf photosynthesis. More important isth eeconomi cdamag edu et oresidue s of stickyhoneyde w on fruits, ornamentals andcotto nlin t(Byrn e& Bellows , 1991;Schuste ret al, 1996).Sometime s themer epresenc e of only a few whitefly nymphs and adults can cause problems. 3) The marketability of ornamental plants depends on aesthetic features. The presence of whitefly adults and honeydew will not escape the notice of consumers. For export of plants orplan t products a zero-tolerancei svalid ,especiall ywhe nth einsec ti sa quarantin epes ti nth eimportin gcountry . Infested plants destined for the international market risk rejection. Interception can lead to destruction orre-expor t of anentir e shipment (Fransen, 1993;Schuste r etal., 1996).4) Both Chapter 1 whitefly species servea svector so f severaleconomicall y important viral plant pathogens (for overview seee.g. Brown etal., 1996 ;Duffus , 1996).Hence ,onl ya fe w adults carryinga viru s can cause serious damage to acrop .
Whitefly biology Adultwhiteflie s arewinged , free-flying insects.Th efemale s laythei r eggspreferabl y onth e abaxialsid eo fyoun ghost-plan tleave s(va nLentere n& Noldus , 1990).Thei regg sar einserte d intoth elea f tissueb ymean s of apedice l (Paulson &Beardsley , 1985).Throug h thispedice l watero f theplan tca nb eabsorbe d toovercom e desiccation of theegg s (Byrneet al., 1990b). Thehatchin gcrawler s (first-instar nymphs)usuall y settlei n theproximit yo f the oviposition site following a search for a suitable location and start feeding. Sometimes the crawler can coverlarg edistances , whichha sbee n attributed tohost-plan t suitability (van de Merendonk &va nLenteren , 1978). Onrar eoccasion scrawler scoul deve nmov ebetwee nplant s(Summer s etal, 1996).A ssoo n asa suitabl elocatio n isfoun d theybecom e sessileb yfixin g themselves to the leaf with their mouthparts. All four nymphal instars, following the settling of the crawler, will spend their life at the same location, reinserting their mouthparts into the leaf tissuefollowin g eachmolt .Th enympha lstage sca nb edistinguishe db ysevera l morphological characters andeasies tb ysiz e(Weber , 1931;Lopez-Avila , 1986).Th efourt h instar transforms into a so-called pupal stage during which the scale-like nymph transforms into the winged adult that emerges through a sliti n thedorsu m of its shed exoskeleton. Theduratio n of this life-cycle isdependen to nwhitefl y species,temperatur ean dhost-plan tsuitability .Fo rinstance , thedevelopmenta l time ofT. vaporariorum onpoinsetti a isfaste r than that of B. argentifolii at20°C ,bu ta t2 5° Ci ti sth eothe rwa yroun d (Fransen, 1994).Fo rmor edetaile d information about whitefly morphology, life-history parameters, behaviour or host-plant preference see, for example,va nLentere n &Noldu s (1990),Byrn e &Bellow s (1991),va n Roermund &va n Lenteren (1992)o rGerlin g(2000) .
Whitefly management strategies In natural ecosystems and agro-ecosystems where nopesticide s areused , an arrayo f natural enemies can keep whitefly at low numbers by a combination of predators, parasitoids and pathogens.Severa lexample sar eknow no fperfec t naturalcontro lo fwhitefly , e.g.i ntomatoe s inth e 1960'si nCalifornia , incotto n duringth eperio d 1925- 1992i n Sudan (vanLentere net al., 1996), in citrus beginning 1900's in Florida (McCoy, 1985) or in pimento on Jamaica (Borner,1956) .Whe ninsecticide san dfungicide s areapplie dnatura lenemie sar eexterminate d and whitefly pests are theresul t (van Lenteren etal., 1996) . General introduction
In protected crops, especially in ornamentals, the low tolerance of pest insects is reflected bywidesprea d and frequent useo f insecticides.Biologica l control infloricultur e is moredifficul t compared withbiologica lcontro l infrui t vegetablesfo r severalreasons .Firstly , more chemicals are available for ornamentals than for fruit vegetables, because of safety regulations for consumption of vegetables. Secondly, in fruit vegetables only the fruits are harvested, which allows ahighe r population level of thepes t onth eleaves . Incontrast , most ornamentals arebein g soldwit hflower s andleaves ,an dtherefor e shouldb efre e of injury and presenceo finsects .Thirdly , the criterion of zerotoleranc e for both pest and natural enemies hasbee nuse d asa standar d for allproduct s until now,thoug h iti s onlyneede d asa nexpor t requirement for a few countries (Fransen, 1993). Nowadays, increasing knowledge and concern about the effects of chemical pesticides on the environment and the problem of resistence development against pesticides by insects (Dittrich & Ernst, 1990; Horowitz & Ishaaya, 1996)ma ycaus echemical st ob en olonge ravailabl ei nth efuture . Therefore research has shifted towards other means of control, such as integrated pest management (IPM), includingbiologica l control andguide d control, andbreedin g for host-plant resistance, using chemical control only as an ultimate resort. Biological control using predators, parasitoids and/or pathogens, can offer areliabl e method tocontro l whitefly.
Predators and parasitoids About 75 species of whitefly predators havebee n described. However, many more species preyupo nwhiteflies , likegenera l predators such as spiders,beetle s etc.(va nLentere n et al., 1996).Sinc ewhiteflie s areminut einsects ,man yo fthei rpredators ,i nparticula rth eone smor e specific tothei rprey ,ar esmal larthropods .Th eposition so fth ewhitefl y eggs andnymph so n theundersid e ofleave sals olimit sth ekin do fpredator s tothos etha toccu ro nleave s(Gerling , 1990). Individual predator species in the families of Anthocoridae, Coccinellidae, Chrysopidae, Hemerobiidae and most of the Miridae are unable to maintain e.g. the greenhouse whitefly belowdamagin glevels ,althoug ha comple x ofpredator sma yd oso .Onl y some of the predatory mirid species belonging to the generaMacrolophus orDicyphus may be able toreduc e whitefly populations sufficiently. Their polyphagy allows them to feed on otherprey ,lik emites ,aphids ,thrip s ornoctui d eggsan d hencemaintai n their numbers when whitefly densities arelo w (Onillon, 1990;va n Lenteren etal., 1996) . Circa 100 species of whitefly parasitoids have been identified and more species are expected to be found. Most of the parasitoids are very host specific, but some species are hyperparasitoids and their importation might reduce theefficac y of primary parasitoids (van Lenterenet al., 1996) .Man yimportan twhitefl y parasitoidsbelon gt oth egener aEncarsia an d Eretmocerus (Hymenoptera: Aphelinidae), and the genus Amitus (Hymenoptera: Chapter 1
Platygasteridae), for example, Encarsiaformosa, Eretmoceruscalifornicus, E. mundusan d A. fuscipennis (Gerling, 1990; Fransen, 1994; Manzano, 2000). Biological control of T. vaporariorum with E.formosa is one of the success stories in biological control, andth e parasitoid is aver yreliabl e control agent in crops as tomato, sweet pepper and gherkin,bu t less so in egg plant and cucumber. In ornamentals, such as gerbera, results are ambiguous, dependingmainl yo ntemperatur e (Roermund, 1995; vanLentere n &Martin , 1999;Siitterlin , 2000).
Pathogens In general, insect pathogensbelon g to verydivers e taxonomic groupslik e viruses,bacteria , protozoa, rickettsiae, fungi and entomophagous nematodes. For whiteflies the spectrum of pathogensi smor enarrow .Ther ear en orecord so fnematode sparasitizin gwhiteflie s andwhil e whiteflies mayb ekille db yviruse s orbacteria ,thi s will mainlyb edu et o secondary infection throughalread yexistin gwounds .S ofar , theonl ypathogen s reported from Aleyrodidae have beenfung i (Fransen, 1990).I ngenera lth efung i infecting Aleyrodidae canb edivide dint otw o groups: thosespecifi ct oAleyrodidae ,mainl ybelongin gt oth egenu sAschersonia (teleomorph : Hypocrella), andth e"broa d spectrum fungi" (Fransen, 1990).Funga lspecie sbelongin gt oth e lattergroup ,suc ha sPaecilomyces spp .an dBeauveria bassiana, areals oabl et oinfec t insects belongingt oothe rinsec torders ,an dsometime sthe yar eabl et ohyperparasitiz e onothe r fungi, likefo rinstanc eVerticillium lecanii (Zoub a& Kahn , 1992;Verhaa ret ah, 1998).A novervie w offunga l speciescapabl eo finfectin g different whitefly speciesi nnatur eo ra sbiocontro lagen t is giveni nTab . 1.1 and 1.2. Themod eo finfectio n of variousentomopathogeni c fungi isbasicall yquit e similar.A typical infection cycle involves several steps: conidial attachment, germination, penetration through the insect cuticle, vegetative growth within the host, fungal protrusion outside the insect and conidiogenesis (McCoy et al., 1988). During attachment to the host's body, mucilage associated with conidia and enzymes play an important role. For the subsequent germination of spores ahig hhumidit y ismandatory . The next step,penetratio n of theinsec t cuticle, which rarely occurs via wounds, is thought to be a combination of mechanical and enzymaticmeans ,i nwhic hproteases , lipases and chitinases are the most important enzymes (Butt, 1990).Afte r penetration fungi proliferate within thebod y of thehost . The insectma y be killed by some combination of mechanical damage produced by the fungus, nutrient exhaustion andth eactio n offunga l toxins (Gillespie &Claydon , 1989;But t &Goettel ,2000) . Later, hyphae will emerge from the cadaver, dependent onhumidity , and produce sporeso n the exterior of thehos t (Roberts &Humber , 1981). General introduction
Approaches for microbial control Several approacheshav eevolve d for theus e of pathogens and terminology has been adopted primarilyfro m biological control withpredator s andparasitoids .Thes einclud e 1)permanen t introduction, 2) conservation and 3) augmentation (inoculative and inundative releases) (Hajek, 1993). Permanent introduction is the establishment of an organism in a pest population where it does not naturally occur, which results in more or less permanent suppression (Fuxa, 1987).Environmenta l manipulation orconservatio ninvolve sth enaturall y occurringpes tcontro lb ymean sothe rtha ndirec tadditio nt oth epathoge nunit salread ypresen t (Fuxa, 1987).However , this control strategy has seldom been utilized, possibly because the epizootiology of host/pathogen systems is too poorly understood to determine likely applications (Hajek, 1993).Augmentativ erelease so ffunga l pathogens generally arebase do n the ability of the fungal pathogen to repeatedly cycle through the host population after introduction.Th edegre ean dspee dwit hwhic hfung i areabl et oincreas ei nprevalenc e aswel l as the damage thresholds for the crop determine whether releases should be inoculative (seasonal introduction) or inundative. The latter is frequently described as the use of mycoinsecticides. Among the major strategies of control, inundative releases have been adoptedmos textensively ,especiall yamon gth ehyphomycete s (Hajek, 1993),frequentl y using application technology similar to that used for standard chemical pesticides (Hajek & St.- Leger, 1994).
Table 1.1:Record s of entomopathogenicfung i on whitefly (revised after Fransen, 1990), their origin and status (n = fungus naturally occurring or i= fungus introduced). Fungal species Aleyrodid species Status Country Literature reference ZYGOMYCOTA Zygomycetes: Entomophthorales Conidiobolus coronatus Bemisia tabaci n Israel Gindin & Ben-Ze'ev, 1994 Conidiobolus sp. Bemisia tabaci n Israel Gindin & Ben-Ze'ev, 1994 Erynia radicans Bemisia tabaci n Israel Ben-Ze'ev etah, 1988 Orthomyces aleyrodis Trialeurodes abutilonea n Alabama Steinkraus etah, 1998
ASCOMYCOTA Pyrenomycetes: Sphaeriales Hypocrella spp. seeTab . 1.2
DEUTEROMYCOTA / FUNGI IMPERFECTI Coelomycetes: Sphaeropsidales Aschersonia spp. see Tab. 1.2 Hyphomycetes: Moniliales Acremonium sp. Bemisia tabaci n Danmark Steenberg &Humber , 1999 Trialeurodes vaporariorum i France Riba & Entcheva, 1984 Aphanocladium album Bemisia tabaci n Florida Humber, 1992 Trialeurodes vaporariorum i Bulgaria Entcheva, 1979 Chapter 1
Table 1.1: Continued. Fungal species Aleyrodid species Status Country Literature reference Hyphomycetes: Moniliales Aspergillus sp. Bemisia tabaci n India Humber, 1992 Beauveria bassiana Aleurothrixus floccosus i Spain Santamariaet al., 1998 Bemisia argentifolii i USA Wraightefa/., 1998 Bemisia tabaci n Israel Ben-Ze'ev et al., 1994 Trialeurodes vaporariorum i USSR Borisov & Vinokurova, 1983 Cladosporium aphidis Aleurochiton aceris n Finland Hulden, 1986 Cladosporium herbarum Aleurodicus cocois n Brazil Carvalho et al, 1975 Fusarium scirpi' Dialeurodes sp. n Florida Fawcett, 1944 Fusarium verticillioides Trialeurodes vaporariorum i France Riba & Entcheva, 1984 Fusarium sp. Bemisia tabaci n India Raoetal., 1989 2 n Florida Humber, 1992 Dialeurodes citri n Florida Fawcett, 1908 Dialeurodes citrifolii n Florida Berger, 1921 Metarhizium anisopliae Trialeurodes vaporariorum i Germany Malsam etal., 1998 Paecilomyces cinnamomeus Dialeurodes citri n Florida Fawcett, 1944 Paecilomyces farinosus" Aleurocanthus woglumi n Florida Humber, 1992 Bemisia tabaci n India Nene, 1973 Trialeurodes vaporariorum n Florida Humber, 1992 Paecilomyces Bemisia tabaci n/i Florida Osborne etal., 1990a fumosoroseusA Trialeurodes vaporariorum i USSR Borisov & Vinokurova, 1983 Sporotrichum sp.5 Dialeurodes citri n Florida Fawcett, 1908 Trichothecium roseum Bemisia tabaci n Spain Humber, 1992 Trialeurodes vaporariorum i France Riba & Entcheva, 1984 Verticillium lamellicola 6 Bemisia tabaci n Danmark Steenberg & Humber, 1999 Verticillium psalliotae Bemisia tabaci n Danmark Steenberg & Humber, 1999 Verticillium fusisporum4 Bemisia tabaci i Danmark Steenberg & Humber, 1999 Trialeurodes vaporariorum n Sweden Ekbom& Ahman, 1980 Verticillium lecanit Aleyrodidea sp. n Russia Mor et al., 1996 Bemisia tabaci i USA Meade & Byrne, 1991 n Israel Mor et al., 1996 Trialeurodes vaporariorum n/i UK Hussey, 1958;Hall , 1982 Mycelia sterilia Aegerita webberi Aleurocanthus spiniferus n Taiwan Chien & Chiu, 1986 n/i China Chen etal, 1994 Aleurocanthus woglumi n Cuba Borner, 1956 Aleurothrixus floccosus n South/Central Borner, 1956 America Dialeurodes citri n Florida Fawcett, 1908 Dialeurodes citrifolii n Florida Berger, 1921 Described as F. aleyrodis (Fawcett, 1944). Described as Microcera sp.,nowaday s Fusarium according toHawkswort h etal. (1995). Described as Verticillium cinnamomeum by (Petch, 1932). For geographic distribution see Lacey et al. (1996). Probably Beauveria sp. (Fransen, 1990). Probably a saprophyte (Steenberg & Humber, 1999). General introduction
^5
•c 3
*! 3 "3 TJ- i*l Son SI
00 •ft s3 0« -J 1.3
c o O ON S ""* o a c S •a •8 £S 1 •c ^l o o •as*! on "II Bl « CS "S 3 8 2 'S "O "O "Q c •a 3 ftl •C 'R -C O o 0 1- 0003 sI a O U PQU EuuuEuu J m EEEo c c e c a a a "g ._ a c c a c o •
OT CO *7?
U o a ri,* "a — -is St* CO tfl i; U U s -5 O O ,-r •o y u -a T) 'o "5 .a a i. O O o a •3 '.% § |-a § £ 3 SsJ B So P .is 5 •t •< •< •< «5 CQ CQ Q s 3 -5 J3 .sal 5-3 a»r a II is* o o V a) ii en •£ •S -8 "S °^ i 1 * Chapter1
•a IS S2
"*| OO ft 1 ro •- d\ rl rl (-vJ, .d C\ -H O ON 5 ^ <» Ov Ov
•§2 Eg
•c o O 3 (ii « 3 •fl u rf a T3 •a o i •"§ -§«11§ •c 1 2 .2 3 ,2 2 § § E OH PQ W E oa BH U BH « H? > a c c c •g c a o e
3-S § ^ 3
•o •a •g •O a "a "S I> I5. -3 " '•3 ^•2 '•o a c >: o "•~s js> -Ss> J. S <* £ "3 a Ot? <3 < *QQ
•a
=3
o o J5 OH ffi S J3 OH o S o o B O 2 u Ci o •Si OH OH ^ •s"g a <3 s s £ B I 43 u a 2*O8N So OtN il* •a Ic3 ' 1 r ^ ^~i S 2-a CS ^H M "2 •a TJ Eg •c > H II T3 H 3 (5 •g UJ 0 E •c u O ^<2- DtS U --T3 o4 «I It < 1 7. O 3 BasSis is a a 3 3 0 tH ^- 3 ^H 3 3 -2 O s, III u « m£ « S o < < ft. £ 1 .3 § .„• S ^ cs I" •o "3 •o •o T3 T) •o f. 1-1 S 11 T3 •O T) T) C 6, fr-s « -a X . 00 rt- •* M- J5 ON m *ri in 15 CT\ 0> 0\ 00 V - in ^H ^H -H 00 «ir^- s< c5S •a e J* o) 1 si 5 Z o 18 § s £ S 'So ^ '5b ""1 ^5 J3 •o -O T3 3 Z iia f_ .S, V < Indi a India , cS 5 « pines , .si f o o fGeo r fGeo r Eb 1 s O rf o a .& S 'C o wZea l wZea l wZea l in a 3 .2 S " 0) o a) S o •in ti I" Tl I §"-2 n, o s § Z £ a S d. e ^ III .^r» .ts rr1 ri la o < 111 £ * ^ ! ••* c .2 o u i?l K < < Q fi 3 a. h g •8 PQ £ N U CD s « s si •s* 8* 1a, 10 General introduction >.^ << ~S CM '3 ON ^p r en EC C3 oo 00 X! *U G « *• ON •C «5 "3 J> ft. •* —i a > -J a <*i § « § w W §•" &3 C ass, s o o c/3 C e? 1—1 NO 1 00 offi 0 w x> o 13 ON 00 00 ... ON c.. . OONN ON ."* W3 W O "3 *~' oo •a K •§ ^ rt cs CQ ^ 'O 13 13 •a § - -2 § § I § O <2P sifted § -a ^ ^ 8 3 .2 " "9 :sg , I I .S-8 11 § a -3 la ii QQ Q ; £ 3 S « «) u • & D-i CX I .s M w «i W g 8 e 5-5 g ;g .S. 2 .§ 2 o a) .« *3 ^ cum a 3 S &g rt es cj -ii O S O . •s* ISC i ^ 11 Chapter 1 Abiotic andbioti cfactor s influencing entomopathogenic fungi The success of biological control using entomopathogenic fungi depends on suitable temperature and humidity conditions. Humidity is often cited as the key abiotic factor influencing the potential of fungi. A high relative humidity or free water is important for various parts of the infection cycle i.e. germination, infection and conidiogenesis (McCoy, 1990).Th elowe rlimi to frelativ e humidity for sporegerminatio n isi nth eregio n of 92-93% (Hall& Papierok , 1982).However , infection ofT. vaporariorumb yA .aleyrodis waspossibl e at a relative humidity as low as 50% under controlled climate room conditions (Fransen, 1987). It is thought that the microclimate in the phyllosphere largely determines the performance offungi . Thisma yexplai nth elac ko f correlation between infection byV. lecanii andgreenhous ehumidit y(Ravensber get al., 1990) .Th eobstacl eo fhig hhumidit yrequiremen t canb eovercom ewit himprovement s offormulatio n andapplicatio n strategies(Batema net al, 1993). For fungal growth inside the host and subsequent killing of the insect, the fungus is independent of external conditions. Compared withth erelativ ehumidity , thetemperatur e seems to beo f less importance in glasshouses. Many entomopathogenic fungal species grow and sporulate at temperatures between 15an d3 0° C(Osborn eet al, 1990a).I ngenera lth eoptimu mgrowt han dgerminatio n rateso nartificia l mediavarie d around 25° Cfo rA .aleyrodis an dV. lecanii,an d above 30 °C germination and growth decline rapidly or are impaired (Hall, 1981; Fransen, 1987). The optimum temperature for growth often coincides with optimum temperature for infection. However,temperature sbelo woptimu m notnecessaril ymea n thatinfectio n is less successful. Althoughgerminatio n andgrowt ho ffung i areretarded ,th edevelopmen trat eo fth ehos tinsec t mayb eaffecte d likewise.Thus ,th enumbe r of insects escaping fungal infection bymoultin g willno t increase as aresul t of aslowe r development of the fungus (Hall, 1981). Inadditio nt oabioti cfactors , alsobioti cfactors , sucha shos tplan tan dhos tinsect ,pla y animportan t role in the efficacy of entomopathogenic fungi. Fungal processes takeplac e in theboundar ylaye rsurroundin g thelea f orinsec ti nwhic h theai ri sundisturbed . Thehumidit y in thislaye ri spartl ydetermine db yth ehost-plan t characteristics, such aslea f size and shape, leaf hairiness, position of the leaf on a plant (Ferro & Southwick, 1984). The host-plant chemistry may also influence the fungus directly, for instance, by inhibiting or stimulating fungal germination (Blakeman, 1971) or via the insect, influencing its susceptibility or its development rate(e.g. Hajek etal, 1995;Ellio t etal, 2000;Poprawsk i etal, 2000). Host specificity and virulence of entomopathogenic fungi To evaluate entomopathogenic fungi for the control of insects, several factors have to be considered, such as host specificity, virulence, persistence, epizootic potential and mass 12 General introduction production (Tab. 1.3).Th eimportanc eo fsom eo fthes efactor s dependso nth eapproac hwhic h will be used to introduce the pathogen (Fransen, 1987). For instance, when using entomopathogenic fungi in ainundativ e approach theeconomic s of massproductio n willb e an important factor in selecting a fungal strain. However, when an entomopathogen will beintroduce donc et ocontro l apes tpopulatio n (e.g. permanentintroduction , seasonalrelease) , theepizooti c capacity andsynchronisatio n withit shos t aremor eimportan t criteria (Ferronet al., 1991;Hajek , 1993). How narrow should the host range of an entomopathogenic fungus be, before it is considered for biological control of a pest? By definition, entomopathogens live by parasitizinginsect san di ti suncommo ntha tthes epathogen s alsoattac korganism soutsid eth e Insecta or Acarina (Hajek &Goettel , 2000).On eca n distinguish between physiological and ecological host range. The former is determined in the laboratory under optimal conditions (e.g.hig h relative humidity) and demonstrates which species can be directly infected by the pathogen.Typica leffect s includesom ecombination s of mortality, infection and/or pathogen reproduction. The latter is defined as the host range manifested by pathogens under field conditions (Hajek &Butler , 1999).Knowin g the physiological hostrange , anattemp t canb e made to predict the ecological host range (worst case scenario). However, hosts that can be infected inth elaboratory ,ma yneve rb e found infected inth efield . Thisca nb edu et ocomple x biotic and abiotic interactions in the field, which are difficult to copy in the laboratory. For examplea pathoge n mayb e ablet okil l asub-optima l hostbu tno tb e ablet oreproduc e oni t Table 1.3: Topics for evaluation of entomopathogenic fungi for control of insect pests (after Fransen, 1987). Fungal characteristics: • host range • virulence • sporulation • persistence -epizooti c potential • dispersal • possibilities for mass production • suitability for storage and formulation • toxicological and safety aspects • compatibility with otherpes t and disease control methods Host characteristics: • susceptibility • economic injury level • age distribution, density and spatial distribution in relation to application or introduction strategy Environmental characteristics: • ABIOTIC FACTORS: humidity, temperature, rain, dew, irrigation, wind, solar radiation • BIOTIC FACTORS: host-plant quality, crop structure, leaf sizean d shape, leaf trichomes, microbial interactions 13 Chapter 1 (Hajek & Goettel, 2000). When a strain is specific towards a pest insect it has no direct adversaryeffect s on othernatura l enemies,lik epredator s andparasitoids , orothe r beneficials, likebees . A. aleyrodis, for instance,i sno tonl yunabl et oinfec t theparasitoi d E.formosa, but it is also unable to infect nymphs of T. vaporariorumtha t contain larvae or pupae of E. formosa. In addition, theparasitoi d will reject nymphs infected with the fungus a few days after fungal infection took place (Fransen &va n Lenteren, 1993; 1994).However , being so specific couldb ea disadvantag efo r thegrower , since specific entomopathogens will notkil l other pest insects present in the same crop, which is possible with broad-spectrum entomopathogens such asV. lecanii(Schaa f etal., 1990). Manyspecie so ffunga l entomopathogens arecompose d of genetically diverse groups ofstrains ,whic husuall ycanno tb edifferentiate d bymorphologica lcriteri a(Haje k &St.-Leger , 1994).However ,thes estrain sma yvar ywidel yi nvirulenc etoward sth etarge tinsect s(McCoy , 1990). Animportan tconsideratio ni nmicrobia lcontro lo finsec tpest si sth eselectio no fhighl y virulent strains.Specie so fth egenu sAschersonia, for instance,ar epathogeni c towhitefly , but virulence varies considerably, causing for instance 10% to 92% mortality of greenhouse whitefly (Hirte et al., 1989a). Strains of V. lecanii originating from different hosts show differences in virulence towards T. vaporariorum(Hirt e et ah, 1989b;Masud a & Kikuchi, 1992),althoug hn orelatio ncoul db eestablishe dbetwee nvirulenc eo fth efunga l strainsagains t B. tabaci and their original host (Mor etal., 1996).Repeate d transfer of entomopathogenic fungi on artificial medium can lead to loss of virulence (McCoy, 1990). Virulence can be influenced by nutrition, and sub-optimum conditions of artificial culture can impair normal physiologicaldevelopmen to fth efungu s (Goral,accordin gt oMcCoy , 1990), whereas,passag e through ahos t mayincreas e and/orrecove rth evirulenc e (Hirte etal, 1989a;But t &Goettel , 2000). Thegenu s Aschersonia Fungi belonging to the genera Aschersonia (anamorph) and Hypocrella (teleomorph) are specialized pathogens of whiteflies (Homoptera: Aleyrodidae) and/or scale insects (Homoptera: Coccidae) (Petch, 1921; Evans & Hywel-Jones, 1990). The genera are predominantly(sub-)tropica l andar ecommo nrepresentative so fentomopathogeni c mycobiota innatura l forest ecosystems and plantation tree crops (Evans &Hywel-Jones , 1990; Hywel- Jones, 1993). The geographic distribution of Aschersonia and Hypocrella species, restricted to whitefly, is given in Tab. 1.2 (see also Tab. 1i n Evans & Hywel-Jones, 1990). Very few species are pantropical and there is a clear distinction between Eastern and Western hemispheres. SinceA. aleyrodisan dA. placenta are closely related species (Mains, 1959c; 14 General introduction Samson & Rombach, 1985), the apparently pantropical occurrence might be questionable, although intercontinental movement of host and pathogen on citrus material is a distinct possibility (Evans & Hywel-Jones, 1990;e.g. Oho, 1967an dPonomarenk o etal., 1975).Th e genera are easily recognised by their typically brightly coloured stroma and they can cause spectacular epizootics, implicating them as potent natural control agents of coccids and whiteflies (Evans& Hywel-Jones , 1990).O fth e2 4Aschersonia speciestha t aredescribe d on whitefly, A. aleyrodis andA. placentaar emos tcommonl yobserve d (Fransen, 1990;Tab .1.2) . Thegenu sAschersonia isdivide dint o specieswhic h arerestricte d toAleyrodida e and speciesrestricte dt oCoccida e(Petch , 1921).Subgener awer epropose dbase do nth ehos tinsec t and morphological characteristics, viz.specie s restricted to Aleyrodidae having paraphyses within the conidiomata, whereas species restricted to Coccidae are lacking these structures (Petch, 1921;Evan s& Hywel-Jones , 1990). However,Dingle y(1954 )questione dthi sdivisio n in subgenera, sinceA. duplex,wit h paraphyses init s conidiomata, is consistently associated with Coccidae (see also Tab. 1.2 and Evans & Hywel-Jones, 1990). In addition, most Aschersonia descriptions arebase d onherbariu m material only (Samson, 1995) andi n many cases host identifications are lacking (Petch, 1921;1925 ; 1932; 1939;Mains , 1959a).Hos t identification was impossible when no living hostremaine d and thecomplet e destruction of infected hosts (Evans &Hywel-Jones , 1990).Hywel-Jone sdemonstrate d thatpur eculture so f Aschersonia can lack paraphyses whereas these were present on the host, indicating that paraphyses are apparently not aver yreliabl e taxonomic feature. The genus isi nurgen t need of revision and probably many more species remain to be described from tropical forest ecosystems (Evans &Hywel-Jones , 1997) Theutilisatio n ofAschersonia spp.a sbiocontro l agentsha s alon ghistory .I nth eearl y 1900's,A. aleyrodisan dA. goldiana were mass-produced on sweetpotat o agar and sold to citrusgrower sfo r 75cent spe rcultur efo r thecontro lo fcitru swhitefly , Dialeurodes citri, and cloudywingedwhitefly , D.citrifolii (Berger , 1920b;1920a ; 1921).Toda ythi si s aexampl eo f successful biological control, where predators, parasitoids and fungi control these pests (McCoy, 1985).Simila rresult shav ebee nobtaine di nJamaica ,wher eA. aleyrodissuccessfull y controlled guavawhitefly , Metaleurodicus cardini, (Borner, 1956)an dth eAzerbaija n region, wheresevera lspecie s ofAschersonia hav ebee nintroduce dfro m Southeast Asiaan dAmeric a for the biological control of citrus whitefly (McCoy et al., 1988). Under favourable environmental conditions approximately 80% of nymphal mortality was obtained and the introduced fungus adapted well to new citrus plantations (McCoy et al., 1988). As a consequence, Aschersonia spp. were tested against T. vaporariorum and in protected cultivation in Europe and China as the only regulatory agent or as a component of an1P M 15 Chapter 1 programi ncombinatio n withth eparasitoi d E.formosa (Adam, 1978;Solovei , 1981;Fan get al, 1983;Ramaker s& Samson , 1984;Fransen , 1987;Hirt eetal, 1989a). Aiman d outline ofthi sthesi s Asmor eenvironmentall y responsible cropprotectio n strategies are adopted, natural enemies of both whitefly species, T.vaporariorum and B. argentifolii,wil l play an increasing role. Screeningfo rnatura lenemie swhic har eabl et okil lbot hpes tinsect squickly ,withou t affecting othernatura lenemies ,i sa nimportan tlin eo fresearch .Entomopathogeni c fungi canmee tthes e requirements andca ntherefor e bea valuabl easse tt oexistin gbiologica l andchemica l control measures (Lacey et al., 1996). Our attention is directed towards the genus Aschersonia, becausepreviou sresearc hindicate dtha tAschersonia aleyrodis i sa promisin gwhitefl y control agent because of its high virulence towards greenhouse whitefly, its tolerance to relative humiditiesa slo wa s50 %(Fransen , 1987),it slon gpersistenc eo nlea fsurface s (Fransen, 1995) and its compatibility with E.formosa (Fransen & van Lenteren, 1993; 1994). This project consisted oftw ocomponents : 1)th eidentificatio n of virulentisolate s ofAschersonia spp .fo r theus e againstgreenhous e andsilverlea f whitefly, and2 )th e studyo f factors which influence theeffectivit y ofentomopathogeni c fungi, with special reference to hostplant , humidityan d their interaction. Inchapte r2 virulenc e of over4 0isolate s ofAschersonia spp .agains t B.argentifolii and T. vaporariorum is described. Important aspects for microbial control like spore production and correlation between virulence against B. argentifolii and virulence against T. vaporariorum havebee n studied. Six isolates are selected for further research onbase so f sporulation,virulenc ean dgeographica lorigin .I nchapte r3 dos ean dtim edependen tmortalit y effects of thesesi x isolates onnymph s ofB. argentifoliian dT. vaporariorum are described. Median lethal dose and median lethal time are determined and the bioassay procedure is discussed. The germination behaviour and infection strategy of Aschersonia spp. on B.argentifolii hav ebee n studiedi nchapte r4 b ymean so fscannin gelectro nmicroscopy .Als o the susceptibility of the various nymphal instars of B. argentifolii and T. vaporariorumt o A. aleyrodis andA. placenta is studied. Thisknowledg e will addt oth e understanding of the adaptation ofAschersonia spp .toward s whitefly species. The success or failure of biological control with entomopathogenic fungi is often related to the humidity at leaf level which would be favourable or disadvantageous for the fungi. This phyllopshere humidity is determined by ambient humidity, as well as characteristics ofth ehos tplan t itself.Chapter s 5t o 8focu s onth erol eo f theambien t climate andth ehos tplan t on efficacy ofentomopathogeni c fungi. 16 General introduction Researchb yFranse n (1995)indicate d that conidia ofA .aleyrodis ca n survive onlea f surfaces of cucumber for at least three weeks and were still able to infect nymphs of T. vaporariorum.However , the extend to which the host plant influences persistence of conidia is still unknown. Thepersistenc e ofA .aleyrodis conidi a was assessed onthre ehos t plants of T. vaporariorum,cucumber , gerbera and poinsettia (chapter 5). Persistence was determined byassessin g (1) germination capacity of conidia and (2)mortalit y of greenhouse whitefly nymphs caused byconidi a after aprolonge d period of exposure of these conidiao n thelea f surface. Therol eo fphyllospher ehumidit yi nth eeffectivit y ofentomopathogeni c fungi against T. vaporariorum and Aphis gossypii is described in chapter 6. The hypothesis that phyllosphere humidity can explain differences in insect mortality duet o entomopathogenic fungi wasteste do nfou r different cropspecies ,viz. cucumber, gerbera,poinsetti a andtomato . We measured different climate andhost-plan t parameters, in order to estimate the humidity in the phyllosphere. By studying more than one insect system, one largely sessile: T. vaporarioruman don emobile :Aphis gossypii, w etrie dt oelucidat eth einteraction sbetwee n the different trophic levels. Theambien tclimat etogethe rwit hth ehos tplan tdetermin eth ephyllosper e humidity. However, what is the result of changes in ambient humidity on the efficacy of entomopathogenic fungi within one host-plant species and is this effect comparable for cucumber, gerberaan dpoinsettia ? These questions wereanswere db y studyingth e influence of ambient humidity onth e efficacy ofA .aleyrodis, A.placenta andV. lecanii(chapte r7) . Within one host-plant species differences in morphology and chemical composition mayexist .Th eornamenta lcro pgerber aexist so fmor etha n 600differen t cultivars with large differences in these characteristics (Krips, 2000). A glasshouse experiment was carried out with two cultivars, which differed in morphological characteristics and probably also in chemical composition. The effect of cultivar on the T. vaporariorum population and the eventual effect onA .aleyrodis an d V.lecanii was studied (chapter8) . Inchapte r9 al l studied aspectso f thistritrophi c system are discussed. In addition the steps which still have to be taken before Aschersonia spp. can be implemented in IPM programmes to control whitefly species. Their advantages and limitations are discussed to answer thequestion : doAschersonia spp .hav e afutur e inbiologica l control? 17 Chapter1 18 2 Virulence ofAschersonia spp .agains t whitefliesBemisia argentifolii and Trialeurodes vaporariorum' Abstract Entomopathogenic fungi of thegenu sAschersonia arespecifi c for whitefly and scaleinsects . Theyca nb euse da sbiologica lcontro lagent s againstsilverlea f whitefly, Bemisia argentifolii andgreenhous ewhitefly , Trialeurodesvaporariorum. Fortyfou r isolates ofAschersonia spp . wereteste d for their abilityt o sporulate and germinate on semi-artificial media andt o infect theinsec thost .Seve n isolates sporulated poorly (lesstha n 5.107conidia/culture )an dte n were not able to infect either of the whitefly species. Several isolates were able to produce capilliconidia. Infection level was not correlated with germination on water agar. After a selectionbase d on sporeproductio n and infection, virulence of 31isolate s was evaluated on thirdinsta rnymph s ofbot hwhitefl y specieso npoinsetti a (Euphorbiapulcherrima). Infectio n levelsvarie dbetwee n 2%t o70% ,an dinfectio n percentages ofB. argentifoliicorrelate d with that of T. vaporariorum.However , mortality was higher for T. vaporariorumtha n for B. argentifolii, as a result of a higher 'mortality by other causes' of greenhouse whitefly on poinsettia.Severa lisolates ,amon gwhic hunidentifie d specieso fAschersonia originatin g from Venezuela and Malaysia, A. aleyrodisfro m Colombia, and A. placenta from India showed consistent results inthei r ability toinfec t both whitefly species. 1 Submitted as Meekes, E.T.M, Fransen, J.J. and Lenteren, J.C. van. Virulence of Aschersonia spp.(Fung iImperfecti ) againstwhiteflie s Bemisiaargentifolii an d Trialeurodes vaporariorum. Chapter2 Introduction Controlo fth egreenhous ewhitefly , Trialeurodesvaporariorum, o nglasshous evegetable swit h the parasitoid Encarsiaformosa is one of the success stories in biological control (van Lenteren et ah, 1996). In contrast to the use of E.formosa in vegetable crops, its use in ornamentals isno t that widely accepted, mainly because presence of afe w nymphs or small amounts of honeydew is considered asunacceptabl e in these crops (van Lenteren &Woets , 1988).Wit hth eaccidenta l introduction of thesilverlea f whitefly Bemisia argentifolii(syn .B. tabaci - biotype B) on poinsettia cuttings in the Netherlands in 1987 (Fransen, 1994), the situation has become more complicated. Chemical control of B. argentifolii is not always effective, as this species has the ability to develop resistance to a wide range of pesticides (Horowitz &Ishaaya , 1996).Biologica l controlo fB. argentifolii withE. formosa is difficult, becausei ti sa les ssuitabl ehos ttha nT. vaporariorum (Szaboet ah, 1993;Hente r &Lenteren , 1996). Moreover, when the two whitefly species occur together, the parasitoid shows a preference for greenhouse whitefly (Boisclair etah, 1990;Szab oet ah, 1993). Because of theproblem s B. argentifoliii scausin gi nbot h indoor and outdoorcrops , therang eo f naturalenemie sbein gconsidere d for use against whitefly hasincrease d overth e last coupleo fyear s (Gerling, 1990;va n Lenteren etah, 1997).I n addition toparasitoid s and predators,a rang eo fentomopathogeni c fungi arefoun d onwhitefl y speciesal love rth eworld . Severalfung i arebein gconsidere d for biological control (e.g. Laceyet ah, 1996).I n ordert o select fungal pathogens for controlling whiteflies it is necessary to select isolates which combine the best possible characteristics for killing the target insects under glasshouse conditions. Relevant characteristics are: i) good mass production features, such as high sporulation onartificia l media, ii) high virulence against target organisms and iii)th e ability to withstand the environment in which the pest is occurring (Prior, 1992;Moor e & Prior, 1993).Severa lfungi , likeBeauveria bassiana, Paecilomycesfumosoroseus andVerticillium lecanii,hav ebee ndevelope dint oproduct sfo rth econtro lo fwhitefl y (Ravensberget ah, 1990; Knauf &Wright , 1994;Bolckman set ah, 1995).Previou sresearc hindicate dtha tals oAscher- sonia aleyrodis Webber is a promising whitefly control agent because of its tolerance to relative humidities as low as 50% (Fransen, 1987), its long persistence on leaf surfaces (Fransen, 1995; Meekes et ah, 2000) and its compatibility with E.formosa (Fransen & Lenteren, 1993;1994) . Along with A. aleyrodis several other species of the genus Aschersonia (Deuteromycotina,Coelomycetes ;teleomorph :Ascomycota ,Hypocrella) havebee n reported onwhitefl y species all overth e world (Petch, 1921;Mains , 1959a,b ;Protsenko , 1959).Th e genusAschersonia i s known to cause severe epizootics in whitefly (Aleyrodidae) and scale insects (Coccidae) in the tropics and subtropics (Evans &Hywel-Jones , 1990).Th e species 20 Virulence ofAschersonia spp .agains t whiteflies attacking whitefly mainlyinfec t thenympha l stage.Althoug h 23 specieshav ebee n described on whiteflies (for an overview see Fransen, 1990),littl e is known about their effectivity in controlling both B. argentifoliian dT. vaporariorum in greenhousecrops . In an attempt to identify more virulent isolates,bioassay s were carried out in which spore production, germination and infection of B. argentifolii and T. vaporariorum,wer e measured. In thispape r wepresen t results of this screeningprocedur e for virulent isolates. Materialsan d Methods Whitefly speciesand plant material Bemisia argentifolii was maintained on poinsettia plants {Euphorbiapulcherrima, cv. Goldfinger) in cages at 23 °Cunde r greenhouse conditions. T.vaporariorum wasreare d on gerbera plants (Gerberajamesonii hybrids) in cages at 21 °Cunde r greenhouse conditions. TheT. vaporariorum populations from gerbera wereno tpreconditione d toth e experimental hostplants ,sinc epre-conditionin g of greenhousewhitefl y toa les s suitable host plant usually doesno t alterth e suitability of that specific hostplan t (Thomas, 1993). Bioassays were carried out on young poinsettia plants with 10-15 leaves, in theirvegetativ ephase .O feac hplan tonl yth etw oyoungest ,full-grown , leaveswer eused .T o obtainthir dinsta rwhitefl y nymphs,5 0t o 55adults ,bot hmal ean dfemale , werepu tint oclip - cageso npoinsetti aleaves .Th eadult swer egive nth eopportunit yt ola yegg sfo r 24hours .Thi s resultedfo rB. argentifoliii n25 9( ±20. 8se) ,9 1( ±6. 3se )an d 188( ±20. 5se )nymph spe rlea f inexperimen t 1A, IBan d3 respectivel y (seebelow) ,an dfo rT. vaporariorumi n 83( ±7. 0se ) and 218 (± 23.7 se) nymphs per leaf in experiment 2 and 3, respectively (see below). Differences in number of nymphs per experiment are due to seasonal influence on whitefly rearingsan dfluctuation s inse xratio .Th edevelopmenta lperio dfro m eggt othir dinsta rnymp h at 25 °C was about 14 days for B. argentifolii and 19 days for T. vaporariorum.Fo r B. argentifolii amixtur eo f second andthir d instar nymphs weretreated , because larval cohorts overlap. Fungal isolates Over4 0isolate sbelongin g toth egenu sAschersonia wer ecollecte d from all overth e world, including isolates from existing collections aswel l as isolates from fresh material (infected whitefly nymphs).I nTabl e 2.1a novervie w isgive no f theorigi n ofth eisolates .Thirty-tw o isolates are not identified to the species level and are referred to as Aschersonia sp.. The majority ofisolate sare ,a sfa r asknown ,multisporal ,excep t for Aa5. Sporulating colonieso n PotatoDextros e Agar (PDA, Difco) wereuse dt o make spore suspensions for inoculation of millet-cultures for mass production of spores. Most isolates were cultured on autoclaved 21 Chapter2 CN CN CN CN CN CN CN m"m mm Wffltn ,m . - CM < < <'3 < < + ++ + + + + + + + ++ + ao «"F l V £ * a T3 2 ..In 03 S^ o S-'B S3 rl S •a V 00 : r l o 0< ^ CNI CN o\ m VI t~- Tf B <1 .vdt- ' 2cN "~: : ?""! O ^ N ri d --' oo F- rj '§• Ha •H-H-H-H-H-H-H-H-H-H •H-H-H-H-H-H-H-H-H-H •H-H+t-H-H-H-H-H O a v> 1 -H m -H d "i ^' ^ od \d rj oi O + + + + -H + + + + + + + ++ + ++ + ++ + + + + -± + R + B B E O O O *- ^ *J ^ - 3 £ 21 •3 *3 -a 'a *3 *3 -3 -3 a o J js 8J i 8- a1 8 •§ "2 ed o s '§ ° O W O •*-< Hi 2 "O *o .2 .S3 " .2 .2 .2 cs a S 3 .2 "^ o £ eo ed ca "2 JO IS 3 .9 <-> ° — — —i — <3 § 11 a S2 .S * Illllll-i-i'S JjSS533'Co'3xiiliiillol OOJ3 » 1)5 J &1 - ~ 'C _o _o .o s &«< s « S £ s s s C to a o a> & •a 2 .25 o S J3 >^^>>>>>^>,>, >,^, H Efl U HI I« "-3 t3 i ~ cdcQcdcdcdcdcdcdcdcd J^JJJ^^J•J 1/I3 muumaumauu ll C/3 •§ ° o < «* uquuuuuuuu U »iJ j j j j j j •J 00 *1 in C cd u '5b 2 •a B- U. o —1 — 90 o c « 00 i—i ^H VO oooooooo^-^inmso \0 VO O O -H •C -S •s W o 22 Virulenceo fAschersonia spp.agains t witeflies ., en a CM —' CM j ^H ^ rt r- cn in ^o ^ o o •* c •^ 00 o ^ O © -^ © o o o © l~ o 'So -H -H -H -H -H -H •H c cMo ^- N£> T~* CS I— V) '§ oo wS ^d 00 ON ON ON 00 00 o> ON ON ON ON ON ON ON ON ON O ^ e o •a + + + 4* + + + + rt + + + a 5 a <;oo oQ . CO too d E E £• ^ CO 0 •5 S 0 0 o o 5 3 U U U i) U Jj 0 0 •s "S o ^ 0 I-H BO ^ « "O T3 "O < 2 < CO e a III CO a 6 13 !& '5b .2 •a "O 'O 'O *o ^3 ^3 '-3 ^ s->': o 9 S S 2 yi d> - ^ u u u O X < ci < < < < Q Q J3 43 VH 11 o 11 •a o 2£ o u < o U Pi Bi «zS^^ •Sco •o 00 ao o a ^ 0> 23 Chapter 2 millet, asa soli d substrate,i n 300m lErlenmeye r flasks (10g mille t with2 5m ldemineralise d water),wherea s afe w isolateswer egrow no nautoclave d cornflou r (Tab.2.1 ;Fransen , 1987). The Erlenmeyer flasks were closed with sterile cotton for aeration. The cultures were incubated at2 5° Can dL16:D 8 (artificial light).Conidi awer eharveste d from threeweek sol d cultures by rinsing them with sterile demineralised water containing 0.05% (v/v) Tween 80 (Merck). Sporulation andgermination Based on the total amount of conidia harvested from these cultures, five categories were distinguished,rangin gfro m nosporulatio n (-)t omor etha n5 x 109(+++ )conidia/cultur e(Tab . 2.1). In the experiments, the suspensions were standardized to 107 conidia/ml by use of a haemocytometer. Todetermin egerminatio ncapacit yo fconidia , 1 mlo fspor esuspensio n (107spores/ml ) was sprayed onto water agar plates (15 g/1agar-agar , Merck) using a Potter spray tower (Burkard Manufacturing, UK). After incubation for 24 hours at 25 °C in artificial light (L16:D8), percentage spore germination was determined by observing 300 spores per plate (twoplate spe rrun ,max .4 runs) .Germinatio n wasrate dwhe nger mtube s- norma lger mtube s aswel l asthos eformin g capilliconidia -exceede d the width of the conidium. Virulence Allisolate swer epasse d throughB. argentifolii onc ean dth ereisolate swer euse dfo rbioassay s onvirulence .Isolate s which could not bereisolate d werediscarde d from thesetests . Fungal virulence wasassesse di nbioassay swit hB. argentifolii(Exps . 1Aan d IB- the experiment wasrepeate d totes tmor eisolates ) andwit hT. vaporariorum (Exp.2) .Onl ya fe w isolates outo f thosetha tdi d notperfor m well onB. argentifoliiwer einclude d in thebioassa y withT. vaporariorum. Anadditiona l seto fAschersonia isolates were tested onbot h whitefly species at the same time (Exp. 3). The isolates that were used in the experiments are mentioned inTab . 2.1;isolate sAa 4an dAp l wereinclude d inever ybioassay .Fo r assessment ofnatura l mortalityals ountreate d nymphs (untreated control) andnymph s treated with0.05 % Tween in demineralised water (application control) were included in all experiments. Per treatment, three plants and twoleave s perplan t wereused , resulting in atleas t 780, 350an d 828nymph s for atreatmen t inExps . 1A , IB, and3 respectively , forB. argentifolii, and for T. vaporariorum in atleas t 329an d 980nymph s pertreatmen t in Exps.2 and 3,respectively . Two ml of spore suspension was sprayed on the underside of aleaf , attached to the plant, using a Potter spray tower. This resulted in approximately 1.36 x 104 conidia/cm2 poinsettialeaf .Afte r evaporation of water,plant s werecovere d withplasti cbag s for 48hour s 24 Virulence ofAschersonia spp .agains t whiteflies to maintain a relative humidity (RH) of 95 - 100%.Plant s were kept in an air-conditioned greenhouse: RH fluctuated between 40 and 80%dependin g on season andphotoperiod , and thetemperatur e waskep t at aconstan t level of 25° C (±1°C) . The final assessment of mortality took place between two to three weeks after application, when allnymph s hadturne dint opupa e and about 30%ha d developed into adult whiteflies.Mortalit yca nb ecause db yth efungu s orma yb edu et oothe rcauses .A nymp hwa s considered infected when it turned opaque white,orang eo rbrown , depending on the fungal isolateused ; anymp hwa s considered 'dead byothe rcauses ' when theinsec t had desiccated (turnedtransparen tbrown ) andn oapparen t causeo f death wasvisible .Eg g andfirs t nymphal stagemortalit yi sexcluded , sincei tha doccurre d before treatment. Overall mortalitywa suse d for analysis. Statisticalanalysis The(fungal ) treatments wererandoml y distributed between theplants .Fo reac h experiment the proportion of dead nymphs of the total nymphspe rplan t was compared between fungal treatments.Dat awer e analyzedusin gGeneralize d LinearModels ,a binomia l distribution and logit link function(Genstat 5 release 3.22, Payne etal., 1987) Subsequent differences were compared with t-tests with total significance level of P= 0.0 5 (RPAIR,PPAI Rprocedures) , in which probability is adjusted to a = 0.000327 (Exp. 1A), a = 0.000198 (Exp. IB), a = 0.000292 (Exp.2 )an da = 0.025fo r effect caused bywhitefl y species or 0.000321 for effect caused byfunga l treatment (Exp.3) . Results Sporulation Ingenera lconidi aof Aschersonia spp.wer eeasil yproduce d onmille tgrain so rcor nflour . For sixisolate sth echoic eo fmille to rcor nmediu mdi dno taffec t the amounto fconidi a produced (Tab. 2.1).Othe r isolatesproduce d conidia more abundantly onmillet , which maypartl yb e duet o anincrease d surface area for growth. However isolates A10,A13 ,A16 ,A2 2an d Atl showeda highe rspor eproductio n oncor nflou r after 3weeks ,despit eth esmalle rsurfac e area. Of the4 4Aschersonia isolates , isolate A21di dno t sporulate at all(- ) and 7 isolates sporulatedpoorl y(±) ,providin gles stha n 5x 107 sporespe rcultur e(Tab .2.1) .Fo rthi sreaso n these isolates were not included in the bioassays. Seven isolates sporulated very well, producingmor etha n5 x 109spores/culture .Thre eisolate sbelonge dt oth especie sA. aleyrodis from Japan and Colombia, onet oA. placenta from India, one toA, insperata from Java and twot o unidentified Aschersoniaisolates ,A 8 from Malaysia and A18fro m Colombia. 25 Chapter 2 Conidialgermination Although most isolates germinated well on water agar (i.e. above 90% germination 24hrs . post inoculation), afe w isolates germinated at alo w or intermediate level, varying between 3.4 and 61.6% (Tab. 2.1). Also on nutrient-rich medium like PDA, germination of these isolates was poor (Meekes, unpublished results). After four to eight days on water agar, conidiao f theisolate s A10,A13 ,A14 ,A19 ,A22 ,Ap l and Atl, formed one to several long, slender,aeria lstructure swit hsecondar yconidi aa tthei rtip s(capilliconidia ) (Tab.2.1) .I nmos t casesals onorma lger mtube swer e seen,usuall y aver ylo wpercentage .Isolat e Apl, which in generalgerminate d well,forme d capillicondia from conidiaa swel l asfro m thetip so f normal germtube so nwate raga r(Fig .2.1) .Formatio n of capilliconidia wasrar ei n isolates A16an d Aa2. N^ v' '-1 Figure 2.1: Formation of capilliconidia in A. placenta, •// after 6day so nWA .Not eth e formation of capilliconidia onth eger mtubes ; 1)primar y conidium,2 )capilliconidia , bar = 8 urn. Virulence Tenisolates ,A10 ,A12 ,A13 ,A16 ,A19 ,A20 , A21, A22,A25 ,an d Atl, did notinfec t either whitefly species or could not be reisolated from diseased insects (Tab. 2.1). These isolates wereexclude d from thebioassay s on virulence. Onpoinsetti amortalit yo fth econtro ltreatment s(untreate dan dTween-treate dnymphs ) differed considerably between B. argentifoliian d T.vaporariorum. For silverleaf whitefly, mortality variedbetwee n 1.9 and 5.8% (Exps. 1 and 3) and for greenhouse whitefly it varied between 6.9 and 15.5% (Exps. 2 and 3). For B. argentifolii, the mortality in the control treatments is not significantly different from the percentage 'mortality by other causes' 26 Virulence ofAschersonia spp. against whiteflies 70 "1 60 CD CD 40 (D (!) 30 O O O O i3 .Q .Q .O 20 | (0 (0 (0 CO y 10 nj ffl * j * » * » : CO 0 72±:?;LZ ; ._B_.I_I_ I _I_ o C Tw A2 A27 A11 A7 A6 A4 Ai2 Ai1 A23 A8 A9 A17 A14 Aa2 A24 A15 A18 A5 A1 Ap1 A26 Aa4 E 70 >. B CD 60 50 40 30 20 10 CO COtOcOCtJcGTOCQCOCO 0 C Tw A2 A27 A11 A7 A6 A4 Ai2 AM A23 A8 A9 A17 A14 Aa2 A24 A15 A18 A5 A1 Ap1 A26 Aa4 Aschersonia isolates Figure2.2 :Mea npercentag emortalit yo fthir dnympha linstar so fB. argentifolii b yAschersonia spp .(dar kgre y bars)an dothe rcause s(shade dbars) ,A : Exp.1A , B:Exp . IB;erro rbar srepresen t standarddeviation .Bar sno t followed byth esam elette rar esignificanti ydifferen t (1A :a = 0.000198 ; IB:a =0.000327) . observed in the treatments with Aschersonia. Mortality by other causes varied more with T. vaporariorum than with B. argentifolii. For isolates A8, Aa2 and Apl (Fig. 2.3), and A28, Ap2, Aa6 and A32 (Fig. 2.4), mortality by other causes was significantly higher (p < 0.05) than in the control treatments. This could be due to an effect of poinsettia on T. vaporariorum or to overlap between natural mortality and mortality caused by the fungus. For this reason the overall to compare the performance of the isolates. Virulence of theAschersonia spp. for both silverleaf and greenhouse whitefly differed considerably between isolates (Fig. 2.2 - 2.4). In the experiments conducted with B. argentifolii, isolates Aa4 (Fig. 2.2A), A17 (Fig. 2.2B), A26 and Apl (Fig. 2.2A and 2.2B) 27 Chapter 2 100 90 80 vp t^i ra A) & en 60 r o 50 (D E< 40 0> x: 30 S 20 f^i 10 ! i 0 C Tw A27 A5 A14 AM A1 Ai2 A23 Aa2 A28 Aa4 A17 Apl A15 A26 Aa1 AS A24 Aschersonia isolates Figure23 : Meanpercentag e mortalityo f thirdnympha linstar so fT. vaporariorum(Exp .2 )b yAschersonia spp. (light greybars ) and other causes (shaded bars); error bars represent standard deviation. Bars not followed by the same letter are significantly different (a = 0.000292). ^ 70 =5 6° O 50 jg- 40 a3 ig 30 s 20 10 0 C Tw A32 A29 Ap2 Aa4 A30 Aal A28 A31 Apl Aa5 Aschersonia isolates Figure 2.4: Mean percentage mortality of third nymphal instars of B. argentifolii (dark grey bars) and T. vaporariorum (light grey bars) by Aschersonia spp. and other causes (shaded bars) (Exp. 3); error bars represent standard deviation. Bars not followed by the same letter are significantly different (a = 0.000321). caused a significantly higher mortality than the majority of the other Aschersonia isolates tested.Severa lisolates ,suc ha sAl , A17,A1 8an dAa 4di dno tperfor m consistently (Fig.2.2 A vs.2.2B) .O nT. vaporariorum(Fig .2.3 )man yisolate sperforme d well.I nExp .3 a nadditiona l seto fisolate swa steste do nbot h whitefly species (Fig2.4) .Interactio n between whitefly and 28 Virulence ofAschersonia spp.agains t whiteflies fungal treatments was not significantly different (p = 0.66). Overall mortality of T. vaporariorum nymphs was significantly higher than overall mortality of B. argentifolii nymphs. However,thi seffec t ismainl ydu et odifference s inmortalit yi nth econtro ltreatment s or 'mortalityb yothe r causes' between thewhitefl y argentifoliia so nT. vaporariorum. The isolates Aa6, Aa5 and Apl caused a higher mortality on both whitefly species than the majority of the isolates tested in Exp. 3. Four groups of isolates could be distinguished: isolates which were unable to infect either whitefly, those which perform poorly (<10%mortahty),a nintermediat egrou p(10-50 %mortality )an da grou pwhic hperforme d well (>50% mortality). Nocorrelatio n wasfoun dbetwee ngerminatio n levelso nwate raga ran dvirulenc e(Fig . 2.5).Whe ngerminatio no fa nisolat ewa slow ,eve nunde rth ebes tcircumstances ,it svirulenc e wasals olow .However , when germination washigh ,th e isolate couldb e avirulent orhighl y virulent for both whitefly species.Fo r example, germination levels of A14 and Ai2 did not exceed50% ,an dmortalit ylevel sdi dno texcee d25% .O nth eothe rhand ,isolate sA 5an dA1 8 showedhig hgerminatio nlevel s(>92%) ,bu tinfectio n levelsstaye dbelo w 25% (Tab.2.1 ,Fig . 2.2t o2.4) ,wherea sgerminatio n levelso fA2 6o rAa 5als oexceede d 92%,bu tmortalit ylevel s were above50% . 90 I 80 I o T. vaporariorum <$ + B. argentifolii 70 60 O f? ra 50 + °*°* o \ + E 40 0 30 o+ 0 0 + 20 + + \ + + *< Figure2.5 :Mea n percentage 10 + whitefly mortality versus + ° +# + meanpercentag e germination 0 ^ ._*A 40 60 of Aschersonia spp.o n water agar. Germination (%) Discussion Important criteria for development of a mycoinsecticide for biological control are ease of production on a artificial medium, high spore yield and high virulence against target organisms.I nLace yet al. (1996)preliminar y datao nconidia lproductio n ofAschersonia spp. were given. InTabl e 2.1 additional data areprovided . Differences that exist arebase d upon choiceo fmediu m- mille t grains orcor nflou r -,an ddat aove r aprolonge dperio d oftim ear e 29 Chapter 2 included. In general,Aschersonia spp .readil y produce conidia on (semi-)artificial media in a solidphas eo rtwo-phas e system (Oho, 1967;Hirt eet al., 1989a ;Ibrahi met al., 1993 ;Lace y etal, 1996).Artificia l light ordayligh t enhances conidial production (Rombach &Gillespie , 1988). Sporulation ofAschersonia spp.doe s not occur in liquid culture, or, if so, only atit s surface (Ibrahim et al., 1993). This has been a drawback for mass production. However, interesti ntwo-phas ean dsoli dphas efermentatio n technologyo ffung i hasincrease di nrecen t years (Lacey et al., 1996) and it offers new perspectives for mass spore production of Aschersonia spp.i n the future. Amongisolate sof Aschersonia spp.larg edifference s existbetwee ngerminatio n levels on water agar. However there was no clear correlation between germination and infection levels in general. Several isolates were able to form secondary conidia (capilliconidia) on wateragar . Thisphenomeno n was first described forAschersonia b yEvan s (1994),an d iti s morphologically and probably functionally analogous to the capilliconidia of the Entomophthorales.I nth elatter ,the yar ethough tt ob eforme d whennutritiona land/o rphysica l conditions are unsuitable for vegetative growth (Uziel &Kenneth , 1991),fo r instance in the absenceo fa suitabl ehost .Throug hthei rcapacit yt oinoculat e andinfec t mobilehos tinsects , the Entomophthorales are highly specialized to the entomopathogenic 'behavior' (Glare & Chilvers, 1985). Whether capilliconidia in the genus Aschersonia have other virulence characteristics than primary conidia has still to be investigated. Formation of capilliconidia may increase their ability to spread, since primary conidia of Aschersoniaar e produced in mucilage and are mainly dependent on splash dispersal. By formation of one or more capilliconidia dispersal by adult whiteflies or other moving insects to otherplant s would be enhanced. Isolatesof Aschersonia showdifference s invirulenc efo r whitefly. Althoughth egenu s Aschersoniai sspecifi c towhiteflie s (Aleyrodidae) andscal einsect s(Coccidae) ,severa lo fth e isolatestha twer etested ,wer eno tabl et oinfec t B. argentifoliio rT. vaporariorum. Thiscoul d be due to a number of reasons. Firstly, the history of previous attenuation for some of the stored isolates is unknown. This was partly overcome by passage of the isolates through whitefly. For entomopathogenic fungi in general, it has been demonstrated that successive passagethroug h ahos tenhance s thevirulenc e andrepeate d subculturing on artificial medium can decrease the virulence. Changes in the ability to infect were thought to be a result of gradual selection of genotypes (inGoettel , 1992).However , St.-Legeretal. (1991) stated that environmental conditions during growth wereimportant . They showed that levels of cuticle degrading enzymes were higher on conidia that were directly derived from infected hosts compared to conidia produced on artificial medium and, therefore, the former conidia can preadapt toth epathogeni clifestyle .Fo rAschersonia i nparticular ,Franse n (1987) discovered 30 Virulence ofAschersonia spp .agains t whiteflies that, after subculturing A. aleyrodis 12time so n millet, median lethal timeswer elowe rtha n thato fisolate stha twer esubculture d threetimes ,althoug hfina l infection levels were similar. Also Hirte et al. (1989a) stressed the importance of passaging isolates through whitefly in ordert okee pth evirulenc e ata hig hlevel .Afte r 1.5 year, infection levels of theirA. placenta isolatedroppe dfro m 75-95%t o20-40% .Som especie swithi nth eAschersonia genusar ever y difficult to store and to subculture, and viability is soon lost. It cannot be excluded that attenuation of isolates,befor e theycam e inou rpossession , had an effect onth e virulenceo f the isolates inou r experiments. Secondly, the original host of the isolates is not always known. In several cases no livinghos t remained and destruction of the original host byth efungu s was socomplet e that identification wasno tpossibl e (Evans &Hywel-Jones , 1990).Petc h (1921),i n his studyo n thegener aHypocrella andAschersonia, madea clea r distinction between species pathogenic towhiteflie s (Aleyrodiicolae) and species that are ablet o infect scaleinsect s (Lecaniicolae). Morphologically, these two groups differ in presence (Aleyrodiicolae) or absence (Lecaniicolae) of paraphyses inth epycnidiu m(Petch , 1921).Severa l isolates, which areno t describedt ospecie slevel ,ma yhav eoriginate dfro m scaleinsect san dare ,therefore , not able toinfec t whitefly. Forinstance , A. turbinata (Atl)belongin gt oth eLecaniicola edi dno t infect either of the whitefly species. On the other hand, since Petch's monograph of the genus in 1921, several authors described species belonging to the Aleyrodiicolae on scale insects: A. duplex(Dingley , 1954),A. goldiana (Wolcott, 1955), A. aleyrodis (Vargas Sarmientoet al., 1995)an dA . placenta(Li met al., 1991).Li met al. (1991 )isolate dA . placentabelongin g to the Aleyrodiicolae from Asterolaciumungulata (Homoptera : Coccidae) in Malaysia, where thefungu s wasabl et osuppres sthi sscal einsec tconsiderably . Vargas Sarmiento etal. (1995 ) used anA. aleyrodisisolat eoriginatin gfro m Coccidaefo r controlo fgreenhous ewhitefly . The possibility existstha t anumbe r of species likeA. aleyrodisan dA. placenta are less selective to their host as has been presumed thus far. Recently, it has been shown that presence of paraphyses in the pycnidia on the host may lack when the fungus is grown in pure culture. Presenceo rabsenc eo f paraphysesi slikel yt ob ephenotypicall y determined and, therefore, it isno ta reliabl etaxonomi cfeatur e(Evan s& Hywel-Jones , 1997).Th eresult smentione d above alsoindicat etha tspecificit y ofinfectio n within theAleyrodida efamil y does not seem to exist for allspecies .Fo rexampl e anA . placenta isolate originating from Dialeurodes cardamomi (Apl)wa sperformin g equallywel lo nB. argentifolii andT. vaporariorum (Fig.2. 2 and 2.3). Themajorit y ofth eisolate smentione d inTab .2. 1wer eno tidentifie d tospecie s level. Iti sprobabl etha tsevera lo fthe mbelon gt oundescribe d specieso fth egenu sAschersonia. The taxonomyo f thisgenu si suncertai n and urgently innee d of revision (Evans &Hywel-Jones , 1997).T oinvestigat erelationship samon gth eAschersonia speciesan disolates ,Oborni ket al. 31 Chapter2 (1999) reconstructed a phylogeny, using randomly amplified polymorphic DNA (RAPD) markers. In this phylogenetic tree, clustering of isolates coincides with their geographical origin. Isolates A15,A1 7 and A24 probably belong to A. aleyrodis,wherea s isolates A28, A30,A3 1an dA3 2ar eclosel yrelate d to,o rbelon g to, A. placenta. Morphology suggests that themajorit y ofth evirulen t isolates,suc h asA17 ,A24 ,Aal , Aa4,Aa5 ,Aa 6an dAp l belong toth eA. aleyrodis- A. placenta complex. Among the isolates that were able to infect whitefly, a positive relationship existed betweenvirulenc e againstB. argentifolii andvirulenc e againstT. vaporariorum. When testing fungal isolates on both species under the same circumstances (Exp. 3),overal l mortality on greenhouse whitefly is significantly higher than that on the silverleaf whitefly, due to the higher 'mortality by other causes'. Statistical analysis is carried out on data including 'mortalityb yothe rcauses' .Thi smortalit yma yb esimila rt omortalit yoccurrin gi nth econtro l groupo fhosts ,bu t isi nsom ecase slowe ro rhigher .Mos tinsect sar ekille db yth efungu s and asa resul tceas et ob e susceptible toothe rmortalit ycausin g factors (Fransen, 1987).Als o the possibilityo ftoxin splayin ga rol ei nth einfectio n processcanno tb eexclude d (Krasnoff etah, 1996).Fo r this reason, weuse d overall mortality datarathe r than applying Abbot's formula (Abbott, 1925). Poinsettia, with higher developmental rates for B. argentifolii than for T. vaporarioruma t2 5°C ,appear st ob ea mor esuitabl ehos tplan tfo r theforme r whitefly species (van Lenteren & Noldus, 1990;Fransen , 1994). Higher infection levels are to be expected when adevelopmenta l period of the insect isextende d and thehos t is weakened (Steinhaus, 1958).However ,th ehypothesi stha tgreenhous ewhitefly , withit slonge rdevelopmenta lperio d at 25 °C and its higher control mortality, is more susceptible to infection than silverleaf whitefly isno tsupporte db you rresults ,althoug hthi stheor yi sno tbase do ntw oinsec tspecies , but on one species under different circumstances. The mortality levels of B. argentifoliio rT. vaporariorum byAschersonia spp .see m to be lower than mortality levels by other fungi like B. bassiana, M. anisopliae, P.fumosoroseus an dV. lecaniifoun db yothe rauthor s(Vidale fal., 1997;Malsame/a/. , 1998; Wraight etal., 1998;Gindi n etal., 2000).However , oneha s to take into account that most bioassays arecarrie dou t atdifferen t climatesettings ,wit hdifferen t compositions of theinsec t population and/oro ndifferen t host-plant species.Ou rbioassays ,fo r instance,wer ecarrie d out in amor epractice-relate d environment ando npoinsettia , which seemst ohav e amor e hostile environment for entomopathogenicfung i than cucumber andgerber a (Meekes etal., 2000 ;E . Meekes &E .Beerling ,unpubl.) .Thi smake s comparison ofresult so f thevariou s bioassays more difficult. The utilization ofAschersonia species asbiocontro l agentsha s alon ghistory . Inth e early 1900's,A. aleyrodiswa ssuccessfull y introducedi nFlorid afo r thecontro lo fDialeurodes 32 Virulence ofAschersonia spp.agains t whiteflies citrian dD. citrifolii (Berger, 1921).Toda ythi si sa classica l exampleo f successful biological control,wher epredators ,parasitoid s and fungi control thesepests .Simila r resultshav ebee n obtainedi nth eAzerbaija n region,wher esevera lspecie sat Aschersonia hav ebee nintroduce d from South-East Asiaan dAmeric afo r thebiologica l control ofcitru swhitefl y (McCoy etal., 1988).A sa consequence ,severa l specieso fthi sgenu shav ebee n tested as aselectiv e control agentagains tgreenhous e whitefly inprotecte d cultivation in Eastern Europe andChina , with goodresults . Several of theAschersonia isolates tested in our study show goodpotentia l for biological controlo fB. argentifolii andT. vaporariorum through their high spore production on semi-artificial mediaan dhig hlevel s ofinfectio n ofwhitefl y onpoinsettia . Performance of theisolate sA24 ,A26 ,Apl , Aa5an dAa 6wil lb estudie di n moredetai l toge ta bette r insight in(1 )thei rspee do fkil lan d(2 )thei rmortalit yeffect s relatedt oconcentratio n ofspores .Thes e characteristics arecrucia l for obtaining effective biological control. Acknowledgements We are grateful to J. Tolsma for technical support and to J.Th.W.H. Lamers for advice on statistical analysis.W ethan k H.C.Evan s (CABIBioscience ,UK) ,R . Humber (USDA, NY, USA), L.A. Lacey (USDA-ARS, Yakima Agr. Res. Lab, USA), R. van der Pas (Koppert Biological Systems, The Netherlands), H. Saito (Shizuoka Agric. Exp. Stn., Japan), S. Selvakumaran (Indian Cardamom Res. Institute, India) and R.A. Samson (CBS, The Netherlands) for kindly providingAschersonia culture s or infected insects for our research. We also thank the G.J. Bollen (Lab. of Phytopathology, WU, The Netherlands), E.A.M. Beerling and the Entomology PhD group for critically reviewing earlier versions of the manuscript. This research was funded by the Dutch Product Board for Horticulture (Productschap Tuinbouw). 33 Chapter 3 Introduction The greenhouse whitefly, T.vaporariorum, and the silverleaf whitefly, Bemisia argentifolii (B.tabaci biotyp e B) are among the major pests in field and greenhouse crops (Byrne et al., 1990a; Becker et al, 1992). Chemical control has often proven to be ineffective, as these species have the ability to develop resistance to a large range of pesticides (Horowitz & Ishaaya, 1996). Due to the negative side-effects of chemical insecticides and consumer concerns,th epolic yo fman ycountrie sno waim sa ta decreas ei npesticid eus ean d encourages thedevelopmen t of alternativemean so fcontro li n thecontex to fintegrate d pest management (vanLenteren , 2000).On eo f thesemethod s involves theus e of insect pathogenic fungi. The interest for theus eo f thesefung i ishig hbecaus e of, among others,thei r high virulence, their adaptation to and persistence in theinsec t habitat, and their compatibility with other natural enemies (Laceyet al., 1996) .T odevelo p fungal pathogens intomicrobia l control products it isnecessar yt oselec t genotypeswhic h combineth ebes tpossibl echaracteristic s for killing the target insects.I ngeneral , these characteristics canb e divided into i)goo d field performance, showinghig h virulence againsttarge t organisms andth eabilit y to withstand the environment in which thepes t occurs, andii ) good massproductio n features, such ashig h sporulation on artificial media (Prior, 1992). Over3 0differen t specieso f fungi areabl et oinfec t whiteflies, includingAschersonia spp., Verticillium lecanii,Beauveria bassiana an d Paecilomycesfumosoroseus (Hall, 1982; Fransen, 1990;Wraigh tet al., 1998 ,Chapte r 1) andsevera lhav ebee ndevelope dint oproduct s for whitefly control (Ravensberg etal, 1990;Knau f &Wright , 1994;Bolckman s etal, 1995; Wraight & Carruthers, 1999). However, results vary with glasshouse climate, culture conditions ofth ecro pan dcro p species used. Of the abovementione d 30fungi , 24belon g to the genus Aschersonia (Deuteromycotina: Coelomycetes) (Petch, 1921;Mains , 1959a, b). Thesefung i havebee nrecognize dt ocaus enatura l epizooticsi nAleyrodida e inth etropic san d subtropics, andar ecapabl eo fwipin gou t entirewhitefl y populations (Evans &Hywel-Jones , 1990).Th eutilizatio n ofAschersonia spp .a sbiologica l control agents has alon ghistory .A t thebeginnin go fthi scentur ycrud einocul ao fA .aleyrodis andA. goldiana werealread yi nus e tocontro l whitefly populations ofDialeurodes citri andD. citrifoliii n citrus grovesi nFlorid a (Berger, 1921). Today this is a classical example of successful biological control, where predators, parasitoids and fungi control these pests (McCoy, 1985).A. aleyrodis was also tested asa potentia l selective control agent against greenhouse whitefly. Itwa s found tob ea promising whitefly control agentbecaus eo fit stoleranc e torelativ ehumiditie s aslo w as50 % (Fransen, 1987;Chapte r7) ,it slon gpersistenc eo nlea f surfaces (Fransen, 1995;Meeke set al, 2000;Chapte r 5) and its compatibility with theparasitoi d Encarsiaformosa (Fransen &va n Lenteren, 1993;1994) . 36 Dose and timeeffect s ofAschersersonia o n whitefly Previousresearc hha sfocusse d onA. aleyrodis asa contro l agentfo rT. vaporariorum, butwit hth eaccidenta lintroductio n ofB. argentifolii inDutc h glasshouses in 1987 (Fransen, 1994),a biocontro l agentwa sneede dwhic h wouldwor k againstbot h whitefly speciesequall y well.I nchapte r2 i twa sshow ntha tfro m atota lo f4 4Aschersonia isolate steste dsevera lwer e ablet ocaus ehig hlevel so fmortalit yi nbot hT. vaporariorum andB. argentifolii populations. For this study six Aschersonia isolates were selected, based on the quantity of spore production,hig hmortalit ylevel so fbot hwhitefl y speciesan dgeographica l origin (Chapter2) . By taking six isolates from very different regions a wider range of genetic variation was ascertained (Obornik et al., 1999). The virulence of these pathogens was compared by measuring the number of conidia required to kill a certain proportion of a group of insects, expressed as median lethal concentration (LC50, conidia/ml) or median lethal dose (LD50, conidia/cm2). Virulence was also compared by measuring the time required to kill acertai n proportion of a group of insects, expressed as median lethal time (LT50) and recording the nymphal instar in which mortality occurred. Here we report on the results and bioassay procedures for the identification of virulent isolates for further research. Material andMethod s Whitefly host The B. argentifolii culture was maintained on poinsettia plants {Euphorbia pulcherrima 'Goldfinger') in screened cages at23° C under greenhouse conditions. TheT. vaporariorum colony wasreare d on gerbera plants (Gerberajamesonii hybrids) in screened cages at21° C under greenhouse conditions. Toobtai n secondinsta rnymphs ,abou t5 0adults ,bot hmal ean dfemale , were confined toclip-cage s (0 2cm )o nth eundersid e of poinsettia leaves (1clip-cag e perleaf) . The adults were given the opportunity to lay eggs for 24 hours. This resulted for B. argentifolii in approximately 105( ±2. 5SE )nymph spe rlea fi nexperimen t 1 and2 , andfo rT. vaporariorum in 103( ±2. 7 SE) and4 8 (± 1.9 SE)nymph spe r leaf inrespectivel y experiment 1an d 2(se e below). Thedevelopmen t from eggdeposi tt osecon dinsta rnymp hwa sabou t2 5t o2 6day sfo r B. argentifoliian d 17 to 19 days for T. vaporarioruma t 20°C on poinsettia. As a result of overlappingnympha lcohort s amixtur eo ffirst, secon d andthir dinsta rnymph swer e sprayed: first: second : third instar = 35% :55 % : 10%fo r B. argentifoliian d 15% : 80% :5 % for T. vaporariorumi n the first experiment. In the second experiment these percentages were: 10% :50 % :40 %forS . argentifoliian d 5% :60 % :35 %fo rT. vaporariorum, respectively. 37 Chapter 3 Statisticalanalysis Datape rse twer eanalyze d using generalized linear models (Genstat 5releas e 3.22,Payn eet al., 1987)wit hisolat e asindependen t variable.Th eproportio n deadnymph swa sbinomiall y distributed and the logistic-link transformation function was used throughout. An LC50an d standarderro r(base do nFieller' sTheorem )wa sestimate d for eachisolate ,usin gth e Genstat 5 procedure FIELLER. Fromth eoveral l mortalityove rtime ,th emedia nletha ltim ewa scalculated . Asever y leaf is evaluated over time, these mortalities were not independent. A separate regression analysis(generalise d linearmodel ,binomia ldistribution ,logi tlin kfunction ) against timewa s therefore carriedou tfo r eachlea fi nrelatio n toeac h ofth ehighe rdosage s (1x 107an d 1 x 108 conidia/ml).Whe nsignifican t differences occurred,t-test swer ecarrie dou tan dth eprobabilit y level was adjusted toth e number of t-tests performed. Fractions mortalityi nfirs t tothir dinstar , proportionalt otota l whitefly mortality,wer e arc-sin square-root transformed. This transformed mortality was analysed using a three-way ANOVA with whitefly species, isolate and concentration (105 to 108 conidia/ml) as independent variables (Genstat 5 release 3.22, Payne et al., 1987). Subsequent differences were compared with t-tests, in which probability (a) was adjusted to the number of t-tests performed (n) (a/n). Results Medianlethal concentration In the first set of experiments the median lethal concentrations, varying between 1.89 x 106 (A31, T. vaporariorum) and 3.93 x 106 conidia/ml (Aa5, T. vaporariorum), were not significantly different for the three isolates.Thes eLC 50'scorresponde d with ados e of 2.57x 103t o5.3 4 x 103conidia/cm 2poinsetti a leaf,respectivel y (Fig.3. 1A an d B,Tab .3.2) . Probit slopesrange dfro m 0.83 (A31- B. argentifolii; Apl -T. vaporariorum) to 1.32 and 1.44 (Aa5- B. argentifoliian dT. vaporariorum, respectively). Inth e second setth emortalit y caused byisolat e A23 did notexcee d 50%. Inorde r to estimateth eLC 50value ,extrapolatio n outsideth eexperimenta l data would be necessary. The valueitself ,however , wouldb equestionable , since anincreas ei nconcentratio n ledt oa lowe r mortality (see Fig. 3.1C and D). For this reason isolate A23 was excluded from the LC50 calculation.Th elatte rphenomeno n wasals o observed for isolate A24,whic h caused alowe r mortalityo fbot hwhitefl y species at 1 x 108conidia/m lcompare d withth emortalit y at 1 x10 7 conidia/ml,henc eth ehighes t concentration for isolate A24 was also excluded from analysis. 6 TheLC 50value sfo r A24an dA2 6wer eno t significantly different, varyingbetwee n0.5 2 x 10 (A24, T. vaporariorum) and 1.38 x 106 conidia/ml (A24, B. argentifolii), which 40 Dose and time effects ofAschersersonia on whitefly Table 3.2: The average median lethal dose (LD50), 95% lower and upper confidence intervals (CI) for Aschersonia spp. on B. argentifolii andT. vaporariorum. B. argentifolii T. vaporariorum LD50 95% low. & up. CI LD 50 95% low. & up. CI Fungal isolate (x 10' con./cm2) (x 103con./cm 2) (x 103con./cm 2) (x 103conicm 2) set 1 Aschersonia sp. A31 3.12 a* 2.17 - 4.50 2.57 a 1.74 - 3.69 A. aleyrodis Aa5 4.77 a 3.26 - 7.01 5.34 a 3.67 - 7.72 A. placenta Apl 3.39 a 2.43 - 4.75 4.27 a 3.10 - 5.87 set 2 Aschersonia sp. A23 - - - - A24 1.87 a2 1.33 - 2.65 0.71 a 0.35 - 1.42 A26 1.57 a 1.16 - 2.12 1.54 a 0.65 - 3.67 *:number s followed byth e same letters are not significantly different. Statistical analyses were carried out per experiment. Since the same data were used for comparing LD50 and LT50 values, the probability level was adjusted according toth e number of t-test carried out (exp.l:': a = 0.008, exp.2:2: a = 0.025); bold or normal letters indicate different sets of comparisons. *:highes t dose excluded for calculations. B.argentifoli i T.vaporarioru m jf- A31 60 /. Ap1 40 Aa5 20 0 untreated 8 o 10 10" 105 106 107 108 tween S 5 — A26 — A24 A23 concentration (conidia/ml) Figure 3.1: Average total mortality of B. argentifolii (A: set 1,C : set 2) andT. vaporariorum (B: set 1,D : set 2) caused by Aschersonia spp. A31,Aa5 , Apl, A23,A2 4 and A26, applied in concentrations of 104 to 10s conidia/ml, 28 days after treatment. Vertical lines represent standard error of mean. 41 Chapter3 correlated witha dos eo f0.7 1x 103an d 1.87 x 103conidia/cm 2,respectivel y (Tab.3.2) .Probi t slopesvarie d between 0.82 (A24 -B. argentifolii)an d 1.02 (A26- T. vaporariorum). Ingeneral ,contro l mortalities,i.e. untreated andTwee n treated whitefly nymphs,wer e not significantly different (p> 0.05 ) from mortality ofnymph s treated with 104conidia/m l (Fig.3.1 ) orfro m eachother ,henc eusin gTwee n asspreade r didno t affect whitefly mortality. However, mortality ofT. vaporariorumnymph s exceeded themortalit y ofB. argentifolii nymphs: 17.1% (±0. 7SE )versu s3.3 %( +0.3 )fo r set 1 and 16.4%( ±0.7 ) versus4.3 % (±0.3 ) for set2 ,respectively . Sincei twa sfoun d for someisolate stha t with increasing concentration of conidia applied, mortality by other causes decreased, we chose not to correct for control mortality. Medianlethal time Themedia nletha ltim ewa scalculate dfo r concentrations 1 x 107an d 1 x 108conidia/m l (see above).I n the first set of experiments, nymphs of B. argentifoliia swel l asT. vaporariorum treatedwit hisolat eA3 1die d significantly earliertha nnymph streate dwit hAa 5an dAp l (Tab. 7 3.3). For B. argentifoliin o significant difference existed between LT50'sa t1 x 10 (4. 6 - 7.1 days)an d 1 x10 sconidia/m l(4. 6- 6. 6days )fo r allthre eisolates ,wherea sfo rT. vaporariorum 7 8 the LT50a t 1 x 10 (6. 0- 9. 9 days) was significantly higher compared to the LT50a t 1 x 10 conidia/ml (4.5- 6. 4 days) (Tab. 3.3). Table33 : The average median lethal time (LT,0) for Aschersonia spp.o n B. argentifolii and T. vaporariorum at concentration of 107an d 10*conidia/ml . LT5C (days) B. argentifolii T. vaporariorum . main effect main effect Fungal isolate 107 108 fungal isolate 107 10s fungal isolate Set 1 Aschersonia sp. A31 46 46 a*1 60 45 a A. aleyrodis Aa5 71 6.6 b 96 64 b A. placenta Apl 66 59 b 99 54 b main effect cone. A2 A B A Set 2 Aschersonia sp. A23 - - - - A24 65 87 b3 54 56 a A26 50 51 a 53 47 a *: numbers followed byth e same letters are notsignificantl y different. Statistical analyses were carried out per experiment. Since thesam e data were used forcomparin g LD50 and LT50 values, the probability was adjusted according toth e number of t-test carried out(exp.l : u.a = 0.004, lsd = 1.212: a = 0.002, lsd = 1.40; exp.2:3: a = 0.008, lsd= 1.76); bold ornorma l letters and capital orlowercas e letters indicate different sets of comparisons. 42 Dose and timeeffect s ofAschersersonia o n whitefly In the second set of experiments isolate A23 was excluded, because extrapolation outsideth eexperimenta l period( >2 8days )woul db emeaningless .A t2 8day safte r treatment moretha n 90%o fth econtro lnymph sha d already developed into adult whiteflies, which are rarelyinfecte d byAschersonia spp.(Franse net al., 1987) .B. argentifoliinymph streate d with isolateA2 4die d significantly more slowly (6.5an d8. 7 daysfo r 107an d 108,resp. ) compared withnymph so fT. vaporariorum (5.4an d5. 6day sfo r 107an d 108,resp.) ,wherea s nymphso f both whitefly species died equally fast when treated with isolate A26 (5.0 and 5.1 days for B. argentifolii, 4.7 and 5.3 days for T. vaporarioruma t 107an d 108, resp.) (Tab. 3.3). No significant difference in LT50wa s found between applied concentrations for either whitefly. Mortalityin first tothird instar period Another methodfo r comparing speedo fkil li scomparin g stagesi nwhic hmortalit y isoccurs . Forinstance ,whe n nymphsdi ei nthei rthir dinsta rfo r oneisolat e andi nthei rfourt h instar for another, isolateon ekill s itshos tmor equickl y than isolate two,whe nbot h areapplie d on the sameinstar .I nTab .3. 4th ewhitefl y mortalityoccurrin gi nfirs t tothir d instarproportiona l to overall whitefly mortality, is shown for concentration 1 x 105 to 1 x 108 conidia/ml. The mortality at 104 conidia/ml was not significantly different from control treatments and therefore excluded. Inse t 1 forB. argentifolii (Tab .3.4) ,isolat eAa 5wa ssignificantl y slowertha nA3 1 and Apl: 67%versu s 83%an d 84%,respectively , of thetota l mortality occurred infirs t tothir d instar. For T. vaporariorum(Ta b 3.4) isolate A31 was also significantly slower than Apl. Furthermore, thehighe r the concentration, the shorter the infection time. Table 3.4: Percentage mortality occurring in first to third nymphal instar proportional to the total mortality occurring per set of experiments. Interactions: Isolate Concentration Whitefly A31 Aa5 Apl 105 106 107 108 1 set 1 B. argentifolii 83 a* 67 b 84 a 76 BC2 81 BC 76 BC 78 BC T. vaporariorum 74 b 71 b 83 a 66 C 74 BC 78 AB 85 A A23 A24 A26 105 106 107 10s 3 set 2 B. argentifolii 35 b 40 b 64 a 53 A* 49 AB 47 AB 37 B T. vaporariorum 31 b 58 a 60 a 52 AB 46 AB 51 AB 51 AB *:number s followed by thesam eletter s are not significantly different. Statistical analyses were carried out per seto f experiments.Probabilit y levels arecorrecte d for thenumbe r of t-test carried out, lsd's are only valid when data are arcsW(fraction) transformed: ': a = 0.003, lsd = 0.089;2: a = 0.002, lsd = 0.105;3: a = 0.003, lsd = 0.125;": a = 0.002, lsd = 0.147. 43 Chapter 3 In set 2, a significantly higher proportion of B. argentifoliimortalit y caused byA2 6 occurredi nth efirs t tothir d instarcompare dwit hmortalit ycause db yA2 4an dA2 3(avg . 64% versus 35%an d40 % mortality in first to third instar, Tab. 3.4).Thi s meant that isolate A24 andA2 3wer eslowe rtha n A26.I ncontras t toth eresult s in set 1,i t wasfoun d that thehighe r theconcentratio n ofconidia ,th elowe r thepercentag e mortality occurring in the first tothir d instar(Tab .3.4) .Fo r 7". vaporariorum (Tab3.4) ,th epercentag emortalit yoccurrin gi nth e first tothir dinsta r wassignificantl y higherfo r isolatesA2 4 andA2 6i nrelatio n toA2 3 (avg. 58% and 60%versu s 31%).Th e overall mortality caused byisolat e A23di d notexcee d 50%an d only 31% o f thismortalit y occurred in first tothir d instar of both whitefly species. Discussion Median lethaldose Five of the six Aschersonia isolates tested here, were highly infective to nymphs of both whitefly species.Media nletha ldose so fthes efiv eAschersonia isolate s variedbetwee n 0.7x 103 (A24) to 5.3 x 103 conidia/cm2 (Aa5) for both whitefly species. The high virulence confirms ourpreviou sobservation si nwhic hsevera lAschersonia isolate sperforme d verywel l (Chapter 2).Th e findings were comparable with theLD 50o fA .aleyrodis i n the experiments of Fransen (1987)whic h was on average 2.2x 103conidia/cm 2 for second instar nymphso f T.vaporariorum o ncucumbe rplants .Unde r laboratory conditions, the LD50value s observed by Wraight et al. (1998) for numerous isolates of P.fumosoroseus, P.farinosus and B. bassiana ranged from 5.0 x 103 to 1.5 x 104 conidia/cm2 against 15 days old nymphs of B. argentifolii(20° C -27°C) .However , theP. fumosoroseus isolates thatVida l etal. (1997 ) used, seem tob e more virulent against second instarB. argentifoliinymph s (LD50value so f 6.2 x 102- 1.3 x 103conidia/cm 2). When comparing LD50 and LT50 values of different bioassays, one has to take into accounttha tbioassay sar eofte n carriedou ta tdifferen t conditions:relativ ehumidit ylevel san d temperatures mayno t be the same,neithe r may the composition of the insect population or host plant. For instance, Vidal et al. (1997) performed their bioassay on sweet potato at a temperature of 25°C, Wraight et al. (1998) used hibiscus plants at 20 - 27°C, and we used poinsettia at 20°C.Furthermore , differences in LC50ma yb eth eresul t of natural variationi n insectpopulatio n andfungus . LC50's,LD 50'san dLT 50's are statistical estimates on aparticula r seto f datacollecte d ata particula r time andthes e values areneithe r biological constants nor measurements (Robertson etal, 1995),bu tonl yrelativ e indicators for virulence, which can beusefu l in screening for virulent strains. 44 Dose and timeeffect s ofAschersersonia onwhitefl y Bioassayprocedure Many bioassays to test virulence of entomopathogenic fungi against whiteflies have been described, ranging from using fourth instar nymphs detached from the leaf (Landa etal., 1994),lea f discs(Drummon det al., 1987 ;Vida let al., 1997) ,detache d leaves (Wraight et al, 1998), rooted, detached leaves (Lacey et al, 1999)t o intact plants (Fransen et al.,1987 ; Malsamet al., 1998) .Althoug h thebioassa ydescribe d above,conducte d onintac tplant s after Fransenet al. (1987),i smor elaboriou stha na bioassa yperforme d onlea fdisc san dno talway s feasible considering space and costs,man y advantages exist. Firstly, side-effects, sucha sunnatura l ageingo fleave s andcoincidin g extra mortality of whitefly nymphs or speeding up of whitefly development, is avoided (Fransen, pers. comm.).Sinc eal lnympha linstars ,followin g thesettlin go fth e crawlers, will spendthei r life atth esam elocatio n (Byrne& Bellows , 1991),lea f discscanno tb ereplaced .I naddition ,i ti s very difficult tomaintai n leaf discs, especially ofpoinsetti a leaves, fresh forfou r weekso n artificial medium.Th eperio dha st ob ethi slon gi norde rt omak e anaccurat eestimatio no fth e whitefly mortality, sinceoccasionall y signso finfectio n were onlyjus t showingi nth epupa l stage(Franse net al., 1987) .Roote dleave smigh toffe r the sameadvantage ,bu tno tever yhost - plantspecie swil lroo ta seasil ya scabbag e (Lacey etal, 1999).Secondly ,usin g intact plants the leaf, and with it thewhitefl y nymphs, stay in their natural position. Contamination by honeydew andrelate d growth ofsecondar y fungi willno toccu r on theundersid eo fth eleaf , in contrast with leaf discs which arepu t upside down upon artificial medium. Although, placing thecontainer s upside down, will partly prevent contamination. Thirdly, performing experiments onlea f discs, instead ofo n whole plants,fo r instance, mayinfluenc e theplan t chemistry andwit hi tinsec tbehaviou r (Risch, 1985;Stam p& Bowers , 1994).I nturn ,th epes t insectca ninfluenc e itshos tplan tsystemicall yan dthroug hthi shav ea neffec t onothe rtrophi c levels,othe rpes tinsect s (Inbaret al., 1999a ; 1999b),predator s (Dicke etal., 1990),bu tals o onentomopathogeni c fungi (Brown etal., 1995).Finally , thebioassay s areconducte d under glasshouse conditions and the results should be easier to extrapolate to large glasshouse applications. Thus, in our opinion, using intact plants in climate conditions related to a practical situation will giveth e mostrealisti cresults . Medianlethal time -first tothird instar mortality The time required tokil l a certain proportion of a group of insects is usually measuredb y monitoringth einsec tmortalit ydaily .Althoug hnymph sinfecte d byAschersonia spp .wil ltur n opaqueorange ,whit eo rbrown ,dependen t onth eAschersonia species,th eexac ttim eo fdeat h isrelativel ydifficul t todetermine ,sinc ewhitefl y nymphsar esedentary .Th emos treliabl ewa y toestimat eth etim eo fdeat h might be recording ofhoneyde w droppings,becaus e honeydew productioni spositivel ycorrelate dwit hnumbe ro fnymph saliv e(Yasu iet al., 1985) .However , 45 Chapter 3 Final remarks Inpresen t strategiest odevelo penvironmentall y safer forms ofpes tcontrol , entomopathogenic fungi willpla ya nincreasin grol ei nth econtro lo fT. vaporarioruman dB. argentifolii. Of the sixAschersonia isolate stested ,isolat eA2 3 wasno teffectiv e becausei tkille dwhitefl y nymphs insufficiently. Isolate A24 was also discarded because, like isolate A23,i t showed a lower mortalitya tth e highest dose and, in addition, spore production of both isolates was nothig h (chapter 2) and their virulence did not compensate for that. Although small differences between isolates obtained in carefully manipulated bioassays mayno t always hold in large- scaleexperiments ,th eabov edat a clearly show thatAschersonia isolates A26,A31 , Aa5an d Apl are virulent against T. vaporariorum as well as B. argentifolii under glasshouse conditions. Their low LD50 and LT50 values strongly indicate that they can be effective biological control agents. Acknowledgements We are grateful for the technical support of N.N. Joosten and J. Tolsma. Special thanks we owe to W. de Boer (Centre for Biometry, NL) and J.Th.W.H. Lamers for their advice on statistical analysis. We thank H.C. Evans (CABI Bioscience, UK), L.A. Lacey (USDA, Yakima Agr. Res. Lab, USA), R. van der Pas (Koppert Biological Systems, NL) and S. Selvakumaran(India n Cardamom ResearchInstitute ,India )fo r kindlyprovidin gAschersonia cultures orinfecte d insectsfo r ourresearch .W e also thank E.A.M. Beerling (PBG, NL),th e Entomology PhDgrou p (WU,NL )fo r criticallyreviewin gearlie r versions of this paper. This research was supported by the Dutch Product Board for Horticulture (Productschap Tuinbouw). 48 Germination and infection behaviour ofAschersonia spp.o n Bemisiaargentifolii and Trialeurodes vaporariorum onpoinsettia 1 Abstract Fungi of the genus Aschersonia are pathogens of whiteflies, which can be used as biocontrol agents against silverleaf whitefly, Bemisiaargentifolii, an d greenhouse whitefly, Trialeurodes vaporariorum.Scannin g electron microscopy and bioassays were carried out to obtain better insight into theinfectio n process ofAschersonia aleyrodis, A.placenta and anunidentifie d Aschersonia sp.. Conidia of Aschersonia spp. germinated readily on the cuticula of host insects as well as on water agar. On water agar A. placenta also produced capilliconidia. No germination was observed on poinsettia leaf surface, except on the leaf veins. On B. argentifolii the fungi formed large amounts of mucilage to attach themselves to the insect. Appressoria were formed before penetration, but also direct penetration was observed. This seemed notrelate d to aspecifi c site on the insect. Forbot h whitefly species, first to third instar nymphs were most susceptible. If thepopulatio n existed of fourth instar nymphs for more than 50%, infection levels dropped from 90 to 50%. Infected whitefly nymphs usually died in the stage following the treated stage. The fungus protruded from the insect viath e margins orvi a the emergence folds of pupae if humidity levels were high enough. However, sporulation was rarely observed on whiteflies developing on poinsettia plants under thepresente d conditions. 1 In preperation as Meekes, E.T.M, Butt, T.M., Joosten, N.N., Porter, R., Beckett, A., Fransen, J.J. and Lenteren, J.C. van. Germination and infection behaviour of Aschersoniaspp .o nBemisia argentifolii an dTrialeurode s vaporariorum on poinsettia Chapter 4 approximately 1.36 x 10 conidia per cm2 poinsettia leaf. As a control, nymphs were sprayed with 2m l of 0.05% Tween 80.Afte r evaporation of the water, plants were covered with plastic bags for 48 hours to maintain a relative humidity (RH) of 95 - 100%. Plants were kept in an air-conditioned greenhouse. The RH fluctuated between 40 and 80% depending on season and photoperiod, and the temperature was kept at a constant level of 20 °C (±1 °C). This treatment procedure was used for microscopic studies as well as the bioassays, with the difference that for the bioassays the leaves stayed attached to the plants until final assessment of nymphs, whereas for the microscopic studies the leaves were picked 24,48 , 72 and 96 hrs after inoculation and processed directly or fixed (see electron microscopy). Microscopic observations Infection processes ofA .aleyrodis, A. placenta andAschersonia A26 were studied in detail on mostly first to third instar nymphs of B. argentifolii. Pieces of poinsettia leaf with infected B. argentifoliinymph s were fixed in 3% (v/v) formaldehyde in 0.05 M phosphate buffer pH 7 for 1.5 h at room temperature (ca. 20 °C). Fixation took place 24, 48, 72 and 96h after inoculation. After fixation, the material was washed twice with phosphate buffer, rinsed with demineralised water and dehydrated through a graded ethanol series culminating in anhydrous acetone (critical point drying, CPD). The leaf pieces with whitefly nymphs were mounted on a copper stub and sputter coated with gold. Specimens wereobserve d with aPhilip s 501B scanning electron microscope. Specimens were also observed by low temperature scanning electron microscopy (LTSEM) in a freeze-dried state with a Philips 50IB scanning electron microscope. Leaf pieces with infected insects were placed on a thin film of water on a copper stub then processed in a Emscope SP2000 Sputter-Cryo Cryogenic-Preparation System using the procedures described byBecket t and Read (1986). Germination of conidia of A. placenta was also observed byfluorescen t microscopy (FM) using the methods described by Butt (1987). In this study 0.02% aqueous Calcofluor (Sigma) was used to stain the fungal cell wall. Specimens were examined with an Olympus BH2photomicroscope , fluorescence images wererecorde d on Kodak TriX Pan400 . Bioassays For assessment of mortality over a range of life stages of B. argentifolii and T.vaporariorum, (fungal) treatments were applied every second day starting after egg hatching (Fig. 4.1). Sixteen treatments in total were carried out on separate sets of plants, one leaf perplant , five leaves per treatment. 52 Germination andinfectio n ofAschersonia on whitefly Since the developmental rate of B.argentifolii nymphs at 20 °C was slower than that ofT. vaporariorum nymphs,ovipositio n of B. argentifoliiwa s scheduled 8day s before the oviposition ofT. vaporariorum. Treatments of both whitefly species were scheduled on the same day. As a result, suspensions of A. aleyrodis, A.placenta and 0.05% Tween solutions (spray control) were applied to the infested poinsettia leaves after 21 to 51 (B.argentifolii) or 13t o4 3(T. vaporariorum) days after oviposition (Fig4.1) . On the day of conidial application, a pre-count was carried out to determine the number of whiteflies present on a leaf and nymphal stage composition of this whitefly population. The latter was carried out to determine the differential infection probability of the nymphal stages. Here, infection probability of the life stages is a combination of 1) contact probability of conidia with their host (hit-probability), 2) the amount of per nymphal stage: nymphal stages differ in size, hence a first instar nymph will receive less conidia than athir d instar nymph, and 3)th e actual susceptibility. For the first sets of treatments, not all whitefly eggs had hatched at the time of pre- count, which could lead to unequal chances of nymphs to come into contact with fungal conidia. This was avoided byremova l of these eggs.Suspension s were applied to poinsettia leaves as described above after 13 to 43 (T. vaporariorum) or 21 to 51 days (B.argentifolii) after oviposition (Fig4.1) . B. argentifolii >"' 0 8 13 21 31 41 51 63 ? / N1 N2 N3 N4 P A T. vaporariorum r ! Ill I r 0 10 13 23 33 434 5 P 2 / N1 N2 N3 N4 A Treatments / / / * -'' 16 /' Figure 4.1:Tim e schedule for treatment of B. argentifolii and T. vaporariorum nymphs at 20 °C: oviposition by female whiteflies (?), hatching of first eggs (S) and the average nymphal stage of whitefly population at the time of treatment (Nl- N4 = first to fourth nymphal stage, P = pupae, A= adults). 53 Chapter 4 A B Figure 4.3: A: Non-germinated conidia of Aschersonia A26 on poinsettia leaf lamina on the base of a minor vein 72 hrs p.i., note remnants of mucilage between conidia (778x, CPD). B: germlings of A. placenta Aplon major leaf vein, 24 hrs p.i. (414x, LTSEM). For all Aschersonia spp. germination on the leaf rarely occurred (Fig. 4.3A), with exception of the major leaf veins (Fig.4.3B ,Tab .4.1) .Th efusifor m conidia (for all species 10-16x 1.5-2.5um ) could form apical, subapical orlatera l germ tubes onth ehos t insect (Fig. 4.4, Tab. 4.1). The majority of Aschersonia conidia on the host insect germinated, exhibiting differential germination behaviour, viz.som e conidia germinated and penetrated directly, others formed appressoria before penetration (within 48 h post inoculation) (Fig. 4.4). Appressorium formation was found on the head-thorax and on the abdomen of the nymphs. On the host insect, the fungus produced mucilage around the conidia and appressoria (Fig. 4.5, Tab. 4.1). Even after the rough CPD treatment, strands of mucilage, attaching the fungus to thehost , were clearly visible (Fig.4.5) .N o preference for a specific site of entrance on the insect was observed. On a few nymphs extensive mycelial growth was observed, without clear evidence of infection (Fig. 4.6, Tab. 4.1). In some cases, the nymph seemed to be able to escape infection by moulting (Fig. 4.7A), but moulting not always results in escape from infection (Fig.4.7B) . 56 Germination andinfectio n ofAschersonia o n whitefly Figure 4.4: Conidia of Aschersonia spp. on B. argentifolii nymphs: Direct penetration of conidia of A. placenta, notice dark zone around penetration site (A: 6250x, 48 h p.i., CPD - B: 3228x, 24 h p.i., LTSEM); C: Appressorium formation by Aschersonia sp. A26 (1543x, 48 h p.i., LTSEM); D: Terminal appressorium of A. placenta (3230x, 24 h p.i., CPD); E: Penetration after appressorium formation of A. placenta, note penetration peg (3230x, 48 h p.i., CPD); F: Apical, subapical and lateral germination by Aschersonia sp. A26 (3230x, 48 h p.i.,CPD) . 57 Chapter 4 E F Figure 4.5: Conidia of Aschersonia spp. on B. argentifolii nymphs: A: Germlings of Aschersonia sp. A26 with mucilage pad (3228x, 24 h p.i., LTSEM); B: Germling of Aschersonia sp. A26 coated with mucilage, despite rough CPD treatment (3230x, 48 h p.i., CPD); C:Mucilag e strands between conidia of Aschersonia sp. A26 (3230x, 48 h p.i., CPD); D: Appressorium with mucilage strands (A. placenta, 1540x, 48 h p.i., CPD); E: Appressoria attached to the insect with mucilage strands (A. placenta, 3230x, 48 h p.i., CPD); F: Germling of with appressorium and maybe a penetration peg, mucilage strands obvious around appressorium (A.placenta, 6280x, 48 h p.i., CPD). 58 Germination andinfectio n ofAschersonia on whitefly Figure 4.6: A: Mycelial growth on abdomen of second instar nymph of B. argentifolii (A.placenta, 788x, 48 h p.i., LTSEM); B: First instar nymph of B. argentifolii with egg shell, both covered with mycelium of Aschersonia sp. A26 (186x, 24 h p.i., LTSEM). Figure 4.7: A: B. argentifolii nymph seems to escape infection by A. placenta by moulting (FM); B: B. argentifolii nymphjus t moulted, without escaping infection byA .placenta (99x, 48 h p.i.,LTSEM) . Four to six days after inoculation, nymphs started to turn orange and when the right conditions were met, mycelial protrusion from the margins and anal opening of nymphs wasobserve d and, depending ondevelopmenta l stageo f thenymph , alsovi ath e emergence folds (Fig. 4.8, Tab. 4.1). On poinsettia, however, sporulation of Aschersonia spp. on whitefly nymphs hardly tookplac eunde r conditions varying between of40-80 %RH . 59 Chapter 4 Figure 4.8: A: Fungal protrusion of Aschersonia sp. A26 from the margins of a second instar nymph of B. argentifolii (advanced stage of infection, 75x, LTSEM); B: Aschersonia sp. A26 ruptured though thorax of fourth instar nymph of B. argentifolii via emergence folds BioassaysB. arsentifolii Mortality of B. argentifolii nymphs by A. aleyrodis and A. placenta varied with nymphal composition of the whitefly population at time of treatment (Fig. 4.9A, B and C). In general, A. aleyrodis and A. placenta caused a significantly higher mortality compared to the control mortality (Tween, on average 3.9% ± 1.8 SE mortality), except when treatment took place after 45 to 51 days for A. aleyrodis and after 49 to 51 days for A.placenta. For both Aschersonia species first to third instar nymphs were most susceptible. When half of the population existed of fourth instar nymphs or pupae (day 37 and further), mortality percentages dropped from > 80% to 50%. This decline in infection probability coincides with the decline of the presence of first to third instar nymphs. First instar nymphs seemed to be slightly less susceptible compared to second and third instar nymphs. However, for the first treatments on average 10% of the nymphs was missing between pre-count and final assessment on the treated as well as on the control leaves. This could be due to crawlers disappearing from theleaf .N o significant differences in mortality existed between A. aleyrodisan dA. placenta. In general, the death of B. argentifolii nymphs did not occur in the nymphal stage which was treated, but one stage later (parallel lines Fig. 4.9D). Since it took on average 7.5 days for B. argentifoliinymph s to develop to the next stage,i t took on average 7.5 days forAschersonia spp .t okil l 70-95%o f the whitefly nymphs at2 0°C . 60 Germination andinfectio n ofAschersonia o n whitefly N3 CD N4 H Pupae H Adults B 21 23 25 272 9 31 33 35 37 39 41 434 5 474 95 1 6? Adult &E P (R O ....._-..-.-- . - - - - - . ^J,r J- U (0 S N4 c a. *"T J*-~ ' 4-^**^ o s N3 5» sis ,,-T sg '•" ^K —e-- pre-count s-g. •o£ N2 < A' —*— A.aleyrodi s >!. » > - • - A.placent a N1 Egg i I I I I I I I I 21 23 25 27 29 31 33 35 37 39 41 43 45 47 495 1 P days after oviposition Figure 4.9: A: composition of the B. argentifolii population at time of treatment (pre-count) shown as percentage first (Nl), second (N2), third (N3), fourth (N4) instar nymphs, pupae or adults of the total population present on a leaf; average whitefly mortality three weeks after pre-count caused by A. aleyrodis (B) orA. placenta (C), error bars represent standard error of total mortality, stacked bars not followed by the same letter are significantly different (rx=0.0000443) , *: not significantly different from control treatment. D: Composition of whitefly population during pre-count and average development stage of the dead B. argentifolii nymphs. 61 Chapter 4 E3 N1 N2 C | N3 C2 N4 Wm Pupae B A. aleyrodis 80 t -o —*! Irk •g T _ ^B Ail -r /' " v ^ r- 1 ^ \i N> NN ^ N V -J7 • ii_SaiMiL 13 15 17 192 1 23 252 7 29 31 33 35 373 9 41 43 A.placent a ililMi.! 13 15 17 192 1 23 25 272 9 31 33 35 373 9 41 43 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 daysafte r oviposition Figure 4.10: A: composition of the T. vaporariorum population at time of treatment (pre-count) shown as percentage first (Nl), second (N2), third (N3), fourth (N4) instar nymphs, pupae or adults of the total population present on a leaf; average whitefly mortality three weeks after pre-count caused by A. aleyrodis (B) orA. placenta (C), error bars represent standard error of total mortality, stacked bars not followed by the same letter are significantly different (a = 0.0000443), *: not significantly different from control treatment. D: Composition of whitefly population during pre-count and average development stage of the dead T. vaporariorum nymphs. 62 Germination andinfectio n ofAschersonia o n whitefly BioassaysT. vayorariorum Mortality of T. vaporariorumnymph s by A. aleyrodis and A. placenta varied also with nymphal composition of the whitefly population at time of treatment (Fig. 4.10A, B and C). From day 13t o 29 after oviposition A. aleyrodisan d A. placenta caused a significantly higher mortality compared toth econtro l mortality (Tween, on average 13.7%± 4. 8 SE).I n this period the majority of the whitefly population existed of first to third instar nymphs. When half of the population existed of fourth instar nymphs or pupae (day 29 and further), mortality percentages dropped from ca. 80% to 50%. This decline in infection probability in the course of the population development coincides with the decline of first to third instar nymphs, as was the case for B. argentifolii. No significant differences in mortality existedbetwee nA. aleyrodisan dA. placenta. In general, death ofT. vaporariorum nymphs also occurred one nymphal stage later than they were treated (parallel lines Fig. 4.10D). Since it took on average 5.5 days for T.vaporariorum nymphs to develop to the next stage,i t took 5.5 days forAschersonia spp . to kill 70-90% of the whitefly nymphs. Although the development of T. vaporariorum population at 20 °C was faster than that of B. argentifolii population, T. vaporariorum nymphs were apparently unable toescap e infection. Discussion Germination In general, these three isolates of Aschersonia spp. show very similar germination and infection strategies. The germination of theAschersonia spp.i s influenced by the substrate. On the leaf lamina conidia rarely germinate. They seem to be either missing germination cues, or germination is inhibited by fungistatic compounds on the leaf surface. However, on the major leaf veins extensive growth was observed. Leaves can leak nutritious phloem substrates particularly near the veins (Isaac, 1992), which may have stimulated conidia of Aschersoniaspp .t o germinate. It could be more advantageous not to germinate in absence of the host, especially since germlings are more sensitive to unfavourable conditions than non-germinating conidia (Sussman, 1968).Conidi a ofA .aleyrodis an dA . placenta seem to prefer to 'sit-and-wait' over growth towards their host. In this way they are able to infect hatching or moulting nymphs of T.vaporariorum even after a month on a leaf surface (Fransen, 1995;Meeke s etal., 2000). Aschersonia placenta is able to produce capilliconidia on water agar (see also chapter 2), but this has never been observed on poinsettia leaf and only occasionally on gerbera leaf in the pocket between leaf lamina and leaf vein. This indicates that high humidity levels are required for production of capilliconidia. Evans (1994) first described 63 Chapter4 this phenomenon for Aschersonia, and it is morphologically and probably functionally analogous to the capilliconidia of the Entomophthorales. Whether capilliconidia in the genusAschersonia have other virulence characteristics than primary conidia has still to be investigated. Formation of capilliconidia mayincreas e their ability to spread. By formation of capilliconidia, dispersal by adult whiteflies or other moving insects to neighbouring plants would be enhanced. Mucilageformation The threeAschersonia isolate s produce huge amounts of mucilage at several stages of their life cycle. Conidia are formed in pycnidia and are covered with mucilage (Samson & McCoy, 1983;Li m et al, 1991;Osborn e & Landa, 1992). Due to the high sugar content the mucus is very hygroscopic (Fransen, unpubl. res.). It facilitates dispersal by free water, like dew and rain, and/or their dispersal byinsect s (Samson &McCoy , 1983). Osborne and Landa (1992) suggested that Acalvolia mites could play an important role in the dissemination of A. aleyrodis conidia and occasional transmission of conidia byEncarsia formosa has been observed (Fransen &va nLenteren , 1993). Due to their mucilage layer, conidia ofAschersonia spp . would not seem subject to wind dispersal (Samson & McCoy, 1983).Nevertheless , Morrill and Back (1912) showed that A. aleyrodis conidia could be spread by wind once they were freed from their mucilaginous matrix after artificial suspension in water (or by rains and dews) and dried. These authors were convinced that wind was the most important factor in long distance dispersal of the fungus from tree to tree and to isolated citrus groves. Rain and dew were considered as animportan t factor for distribution throughout anindividua l treeo rt o closely adjoining trees (Morrill & Back, 1912). Actual establishment of conidia after wind dispersal hasno tye tbee n proven. Conidia borne in mucilage, like those of Aschersoniaspp . and Verticillium lecanii, are hydrophilic and are generally carried in splash droplets (Boucias & Pendland, 1991). The hydrophobic wax layer of insects, which is believed to prevent many fungi to be entomopathogenic, may prevent these conidia from coming into contact with the insect surface. However, conidia of V. lecanii were able to adhere to hydrophobic surfaces, such as the insect cuticle or plastic, in the presence of water (St.-Leger, 1991). Evidence presented aboveindicate s thatthi si s alsoth ecas efo r conidiao fAschersonia spp . Although some remnants of the original conidial matrix will still be attached to the conidium after suspension in water (Fig. 4.3A; Fransen, 1987), once on the host Aschersonia spp. form again large amounts of mucilage around the conidium, to anchor themselves to their host insect (Fig 4.5; Fransen, 1987). These amounts were never 64 Germination andinfectio n ofAschersonia on whitefly observed surrounding conidia on the leaf surface (Fig. 4.3 vs. Fig. 4.5). Also appressoria were surrounded by mucilage strands, clearly visible even after the rough CPD treatment. Mucilage produced by appressorial cells is hygroscopic and may create a favourable environment for the extracellular enzymes released by these structures (Boucias & Pendland, 1991). Infection Not all areas of the insect cuticle are equally vulnerable to fungal penetration (Butt & Goettel, 2000). The intersegmental membranes (Wraight et al., 1990) and areas under the elytra (Butt et al., 1995) can be preferred sites of infection. Butt et al. (1995) showed that germination behaviour of Metarhiziumanisopliae wa s dependent on the cuticle, producing numerous appressoriao nth ecuticl e of afle a beetle, whereas direct penetration occurred on aphids. Here, the presence or absence of germination or appressorium formation in Aschersonia spp. does not seem to be related to a specific site on the insect, nor to a specific instar (mostlyN l toN 3studied) . Occasionally extensive growth of Aschersonia spp. hyphae over B.argentifolii nymphs was observed (Fig. 4.7) without clear evidence of penetration through the cuticle. Whether or not infection was successful, could not be verified, since the SEM procedure is detrimental to the insect. Fransen (1987) did observe this phenomenon too, but on older, fourth instar nymphs of T. vaporariorum. These nymphs were less susceptible to infection, as was confirmed by our results. This aberrant growth of hyphae was also observed for other fungi on resistant hosts, e.g. low pathogenic mutants of Beauveria bassiana showed extensive growth on the surface of Heliothis zeae larvae with only a limited degree of penetration (Pekrul & Grula, 1979). For Aschersonia spp. this might indicate that nymphs of B. argentifolii on poinsettia are less susceptible to infection than nymphs on cucumber (Fig. 9.3; chapter 9), as was shown for T.vaporariorum (Meekes et al., 2000; Chapter 6 and7) . Moulting of nymphs can be considered a favourable and unfavourable factor for fungal infection at the same time. When inoculation takes place just before moulting, the insect mayb e ablet o shedth e fungus together with the old cuticle (Fig. 4.7A). In contrast, the newly moulted cuticle is considered especially vulnerable to fungal attack (Ferron, 1985;McCo y etal, 1988;But t &Goettel , 2000).Although , the insect mayb e able to shed the fungus, there is still apossibilit y it will get infected by settling onAschersonia conidia present onth e leaf surface through itsincrease d size after moulting. 65 Chapter4 Susceptibility First to third instar nymphs were most susceptible to infection by Aschersonia spp. (Fig. 4.9-4.10; Fransen et al., 1987). Some authors report that first instar nymphs of T.vaporariorum o r B. argentifolii were refractory to fungal infection (Hussey, 1958; Hall etal., 1994),other showever , found clear symptoms on first instar nymphs of both whitefly species (Fransen et al., 1987; Meade & Byrne, 1991;Lace y et al., 1999). We also found slightly lower infection levels for first instar nymphs, but infection probability will be lower becauseo f their smaller sizecompare d to second orthir d instarnymphs . As soon as whitefly nymphs enter their fourth instar, they become less susceptible to infection by Aschersonia spp. as already observed by Fransen et al. (1987). The lower susceptibility of fourth instar nymphs and particularly of pupae could be related to the fourth nymphal instar developing into the pupa without moulting. The adult develops inside the pupa, so at the pupal stage the fungus has to penetrate not only the pupal case, but the cuticle of the adult as well.A. aleyrodisi s not able to infect eggs and rarely infects adults of T. vaporariorum(Franse n et al., 1987), in contrast to V.lecanii whic h is able to infect adults, or Paecilomycesfumosoroseus which is also able to infect adults and eggs (Osborne &Landa , 1992). The susceptible instars do not always show signs of infection in the treated stage, but these may appear in subsequent development stages. This phenomenon is frequently observed (Fargues &Rodriguez-Rueda , 1980;Franse n etal., 1987;Vida l etal., 1997) and itca n be used as an indication for the speed of kill for sessile whitefly nymphs (Vidal et al., 1997;chapte r 3).Betwee n 4 to6 day s after inoculation, the infected nymphs turn an orange colour at 20 °C.Whe n the humidity is high enough, the fungus ruptures through the weak spots in the cuticle of the infected insects (Fig. 4.8).Th e fungus-covered host is first white, but it soon turns to an orange-reddish or luteous colour, depending on the Aschersonia spesies,whe n pycnidia are formed and conidiogenesis begins. Waxdeposits On the scanning photographs coiled grooved ribbons (Fig. 4.11) were found all over the leaf and nymphs (e.g.Fig . 4.3,4.5A , E). We wondered whether these particles were host- plant or insect derived. These particles, however, have been described before and were attributed to whitefly adults by Baker & Jeffree (1981), since the same structures were present on, for instance, tomato, tobacco, petunia and dwarf bean leaves and corresponded with presence of T.vaporariorum adults. Adults of both whiteflies produce a lot of wax extrusions, which provide them with their white powdery appearance. The wax particles are excreted from wax glands on the adult's abdomen, broken off by the hind tibia and 66 Germination andinfectio n ofAschersonia on whitefly applied over the whole body and wings. In doing so, these particles are also shed in the environment of the insect. In both whitefly species the wax ribbons have the same configuration, since the templates through which they are forced are similar in shape (Byrne& Hadley , 1988). Figure 4.11: Waxy structures found on first instar nymph of B. argentifolii. These waxes are whitefly (adult) derived (3228x, direct print, LTSEM) Final remarks Although infection byAschersonia spp . is predominately limited to the first three nymphal stages of B. argentifolii and T. vaporariorum, the fungi possess a whole range of germination and infection strategies to infect these stages. Together with their long persistence on leaf surfaces (Fransen, 1995; Meekes et al., 2000), their ability to infect whitefly nymphs in less humid environments (avg. 45%, Meekes, unpubl.) and the compatibility with other natural enemies like the parasitoid Encarsiaformosa (Fransen & van Lenteren, 1993;1994),i tmake s thempoten t biocontrol agents. Acknowledgements We are grateful to J. Tolsma (PBG, NL) for technical support. We thank R. van der Pas (Koppert Biological Systems BV, The Netherlands), S. Selvakumaran (Indian Cardamom Res.Institute ,India ) andL.A .Lace y (USDA-ARS,Yakim aAg . Res.Lab. ,USA ) for kindly providing Aschersonia cultures or infected insects for our research. We also thank the Entomology PhD group (WU, NL) and H.R. Zoomer for critically reviewing earlier versions of the manuscript. This research was funded by the Dutch Product Board for Horticulture (Productschap Tuinbouw). 67 Chapter4 68 Persistence ofth efunga l whitefly pathogen Aschersoniaaleyrodis, onthre edifferen t plantspecies 1 Abstract Persistenceo fAschersonia aleyrodis, afunga l pathogen ofwhitefly , wasstudie d oncucumbe r (Cucumissativus), gerbera (Gerbera jamesonii) and poinsettia {Euphorbia pulcherrima). Germination capacity and infectivity of conidia, which stayed on the different plants over a periodo fu pt oon emonth ,wer eassessed . Average germination of conidia onth eleave s was low (< 14.3%), whereas it was shown that most of the conidia transferred from the leaf to water agar were viable, even after having been present on the leaf surface for one month. Germination capacity was influenced by host plant species: it was highest on cucumber, followed bypoinsetti a andlowes t ongerber a (p< 0.05) .O n cucumber leaves,conidi a stayed viable and were able to infect 90% of the whitefly nymphs, even at 31 days after spore application. On gerbera, germination capacity decreased considerably from 80% (day 0) to 40% (day 31).Thi s was reflected in nymphal mortality, which declined from 75%t o40% . Despiteth ehig hgerminatio n capacity(60% )o f conidiao npoinsetti a after anexposur e of one month, nymphal mortality decreased from 70%a t the day of spore application to 10% after three days at leaf surface, and remained low throughout the monitoring period. Relations betweengerminatio ncapacity ,infectivit y andth ehos tplan tenvironmen t sucha sphyllospher e microbes,secondar yplan t metabolites and microclimate are discussed. 1 Published in slightly different form as Meekes, E.T.M, Voorst, S. van, Joosten, N.N., Fransen,J J . andLenteren ,J.C .van ,2000 .Persistenc e ofwhitefl y pathogenAschersonia aleyrodis,o nthre e different plant species.Mycological Research 104: 1234-1240. Chapter 5 Introduction Asa resul to f lowdamag ethreshold sfo r pestinsect san dmites ,th epresen t controlo f whitefly inornamenta l crops stillrelie slargel yo nchemica l insecticides (Fransen, 1993).I nglasshous e vegetables on the other hand, biological methods to control whitefly are widely used (van Lenteren, 1995). Recognition of negative side-effects of pesticides prompted a shift from exclusive chemical control tointegrate d pestmanagemen t andstimulate d ongoing research in biological control. In the last two decades entomopathogenic fungi, such asVerticillium lecanii, Paecilomyces spp.,Beauveria bassiana an dAschersonia spp. ,hav e shown potential as control agents against greenhouse whitefly, Trialeurodesvaporariorum, an d silverleaf whitefly, Bemisiaargentifolii (e.g. Meade& Byrne , 1991;Lace yet al., 1996) .Unlik ebacteri a and viruses, fungi are able to infect phloem-feeding insects naturally by penetration of the insectcuticle ,whic h makethe m suitablepathogen s of whitefly (Fransen, 1990). Aschersonia aleyrodis causes spectacular epizootics in whitefly populations in the tropics and subtropics (Evans& Hywel-Jones , 1990).Th efungu s wasfirs t usedfo r biological control of whiteflies in citrus groves in Florida in the early 1900's (Berger, 1921), where a combination of predators, parasitoids and A. aleyrodis successfully controlled the citrus whiteflies Dialeurodes citri and D. citrifolii for decades (McCoy, 1985). Aschersonia aleyrodisals o successfully infects the nymphal instars of thegreenhous e whitefly, of which thefirs t threenympha lstage sar emos t susceptible,bu t it isunabl e toinfec t andkil l whitefly eggsan donl ysporadicall y infects adults (Fransen etal., 1987) .Hos t andconidi a maymee ti n two different ways (i) direct contact, when conidia are sprayed upon the insect host or (ii) indirect contact, when hatching or moulting nymphs settle on conidia present on the leaf surface from an earlier treatment. Hence, good coverage and long persistence of conidia on foliage isimportan t for effective control of whitefly. A major difficulty in the development of mycopesticides is their relatively short persistenceo nlea f surfaces.I nmos tstudies ,ultraviole tligh t isconsidere d tob e animportan t factor affecting the survival of propagules (Fargues et al., 1996; Moore et al., 1996). Considering microbial control of whitefly pests in protected cultivation, whitefly nymphs occurmainl yo nth eabaxia l surface ofth eleaf ,wher esurviva lo fconidi ausuall yi shighe rtha n atth euppe r side(e.g. Jameset al., 1995 ;Furlon g &Pell , 1997).I naddition ,glasshouse s filter incomingradiatio n and,therefore , reduceit s (detrimental) impact onth esurviva l of conidia compared with open field situations (Jaques, 1972).Franse n (1995) found that conidia ofA . aleyrodis could survive on leaf surfaces of cucumber for at least three weeks and were still able toinfec t nymphs of greenhouse whitefly. Apartfro m solar radiation, thedirec t environment of conidia may strongly influence their survival. On different host plants the same fungal strain may show differences in 70 Persistence ofAschersonia aleyrodis persistence, as aconsequenc e of chemical and/or morphological differences between plants. Chemicals atth elea f surface caninfluenc e theviabilit y of spores (Cooke &Rayner , 1984), butthe yma yals ohav ea neffec t onth epes torganis man dit ssusceptibilit y topathogen s (Hare & Andreadis, 1983; Ramoska &Todd , 1985).Plan t morphology maydirectl y influence the pestspecie s (BellowsJr .et al., 1994) , butthei r effect onth e microclimate atth e leaf surface, for examplethroug hlea f size,densit yo fhair so nth eleave san dcro parchitectur ema yb emor e important (Ferro &Southwick , 1984;Ingli s etal, 1993). Theexten tt owhic ha hos tplan tinfluence s thesurviva lo fconidi ao f entomopathogens is still unknown. The objective of this study was to assess the persistence of A. aleyrodis conidia on several host plants ofT. vaporariorum. Persistence was determined by assessing (i)germinatio n capacity of conidia and (ii)mortalit y of greenhouse whitefly nymphs caused byconidi a after aprolonge d period of exposure of these conidia onth elea f surface. Materials and Methods Whitefly hostand plant species Experiments were carried out with greenhouse whitefly, Trialeurodes vaporariorum. Whiteflies were reared on gerbera plants (Gerberajarnesonii hybrids) in screened cages at 21°Cunde r glasshouse conditions. Theexperiment s were conducted on three plant species:cucumbe r (Cucumis sativus 'Profito'), gerbera (Gerberajarnesonii hybrids) and poinsettia (Euphorbiapulcherrima 'Goldfinger'). Conidial suspension wereapplie d onlyt ocompletel y unfolded leaves,i n order to avoid alteration in spore coverage, due to growth of leaves. Plants were grown in small glasshouse compartments (30 m2, 21°C, L16:D8) on tables with an ebb and flow irrigation system. Fungus Aschersoniaaleyrodis (KV-107),originatin g from Colombia, was isolated from Aleyrodidae andpassage dthroug hgreenhous e whitefly once.Sporulatin g colonieso npotat odextros e agar (Difco) wereuse dt oprepar ea conidia l suspension for inoculation of millet-cultures for mass production of conidia. The fungus was cultured on autoclaved millet in 300 ml Erlenmeyer flasks (10g mille twit h 25m ldemineralise dwater) ,whic h wereclose d with sterilecotto n for aeration.Culture swer eincubate da t25° Can dL16:D 8(artificia l light).Conidi afro m 2-3wee k old cultures were harvested by rinsing cultures with sterile demineralised water. Conidial densities were determined with ahaemocytomete r and adjusted to 1 x 107conidia/ml . After incubation for 24hr s at 25°C initial germination capacity of conidia on water agar (WA, 15 g/1 agar,Merck ) exceeded95% . 71 Chapter 5 Experimentalsetup Twom lo fconidia lsuspension s werespraye do nth eabaxia l surface ofleaves ,attache d toth e plant, using a Potter spray tower (pressure 3 x 105 Pa, Burkard Manufacturing, UK). This resulted in approximately 1.4 x 104conidia/cm 2. Plants were left to dryfo r onehour , before theywer ecovere d with plasticbag s for 48 hours to create anea r saturated environment(95 - 100% RH). Plants were kept in airconditioned glasshouse compartments (30 m2) at 20 and 25°C (± 2°C); the humidity at 20°C was on average 70%, the humidity at 25°C fluctuated considerably between 40 and 90%,dependin g onth e photoperiod. Germination capacity ofconidia The germination capacity of conidia on leaf surfaces was determined using a leaf imprint technique (Parbery et al., 1981;Fransen , 1995). Clean leaves, attached to the plant, were treated with conidial suspensions as described earlier. On 0, 3, 10, 17,2 4 and 31 days post inoculation (p.i.),lea f discs (0 5cm )wer etake n from inoculated leaves andth e abaxialo fth e leaves was pressed firmly onWA , using aglas s plug of the same diameter. Two leaves per plant and three plants per combination of plant species, time and temperature were used. Germination of conidia on leaf surfaces was determined bymicroscopi c observation of the WA plates directly after the imprint was made. Per plate 300 conidia were evaluated and germination wasassume d whenth elengt h of the germ-tube exceeded the width of thespore . After 24 hours incubation at 25°C, the plates were evaluated a second time, thus providing information on germination of conidia on leaf surfaces and germination capacity of conidia on water agar after being removed from leaves. Theimprin t technique wasverifie d byobservatio n of theconidi a on thelea f surface usinga Zeis sfluorescen t microscope.Conidi ao nleave s were stained with0.01 % Fluorescent Brightener (Merck). The majority of the conidia (80-90%), both germinated as well as ungerminated conidia, were transferred to the water agarb yth e imprint technique used. Whitefly nymphalmortality Fifty whitefly adults, confined to clip cages, were allowed to oviposit for 24 hours on the abaxialsurfac e ofleaves .Thi sprocedur e wascarrie d outever ythre et ofou r dayso n another seto fleaves ,startin g 10day sbefor e inoculationwit hth efungu s until2 1day safte r inoculation (Fig. 5.1); two leaves per plant and five plants per treatment were used. Egg to first instar development ofT. vaporariorumtoo k approximately 10day s at20° C and 8day s at 25°C, so the first set of eggs (day -10)ha d already developed into first instar nymphs at day 0, when application of conidia took place. For the second until the fourth set of eggs, conidia were appliedt oth eeggs ,sinc ehatchin g tookplac eafte r sporetreatmen t atda y0 .Fro mth efift h set 72 Persistence ofAschersonia aleyrodis Inoculation ~r 10 -7 -3 0 4 7 10 14 17 21 24 27 31 1 ? :A 1 -» Daysbetwee ndeposi to f L conidiaan dfirs tpossibl e 2$L :___^_ contactwit hnymph s 32: JL_ 4?L J^_^ „ : 5?L :^_ 6?L ^L 7 2L _!_ 8?L _^_ 10$L_ _^_ _10 Figure 5.1: Time schedule for assessment of persistence of A. aleyrodis conidia by nymphal mortality: egg depositb yfemal e whiteflies (S), eggspresen t on leaf (L- - -) , egghatchin g (•»= first possibl e contact of nymphs withconidia ) andpresenc e of nymphs onlea f untilfina l assessment ( ).Conidi ahav et opersis t from day 0 ( = application) until egg hatching (<*)t o be able to infect T. vaporariorum nymphs (period between vertical dotted lines). ofegg sonwards ,egg swer edeposite do nleave sinoculate do nda y0 wit hA. aleyrodis conidia. About two to three weeks after egg hatching the first adults started to emerge and final assessment for nymphal mortality was carried out. A nymph was considered dead when it turned orangeo r when the insect haddesiccate d andturne d transparent brown. Statisticalanalysis Data were analysed using generalised linear models (Genstat 5 release 3.22, Payne et al., 1987),wit hy representin gth enumbe ro fgerminate dconidi a(germinatio ncapacity )o rnumbe r ofinfecte d nymphs (nymphal mortality) per plant withparameter s n, total amount of conidia or nymphs and p, probability of germination or infection. The proportions of germinated conidia or infected nymphs wasbinomiall y distributed and the logistic-link transformation function wasuse d throughout. Although the effect of temperature should be tested against variation between glasshouses with the same treatment, we were unable to do so, as the experiment was not repeated atglasshous e level (wholeplots) .A n alternative wouldb et operfor m thetes t using restvariatio n ata lowe r stratum.Bu tthi si sa progressiv e test andeffect s aremor elikel y tob e significant. Significant differences weretherefor e treated astrends . 73 Chapter5 Results Germination capacity ofconidia The germination ofconidia on the leaf surface, although low (max. 14.3%),wa s influenced by host plant species (p < 0.001) (triangles in Fig. 5.2). Overall average germination was highest on cucumber (9.8 ± 2.5% se),followe d by gerbera (5.2 ± 1.3%) and was lowest on poinsettia (3.2± 0.7%) .Germinatio n levelstende dt ob ehighe ra t25° C than at20° C for each plant species. After 24hour sincubatio n onwate r agar,th emajorit y of conidia had germinated (Fig. 5.2). Germination capacityofconidi a wasinfluence d byplan t species (p< 0.001) , highest for conidia transferred from cucumber leaves and lowest for conidia originating from gerbera leaves.Germinatio n levelsdecrease di ntime ,wit hth estronges t declinefo r conidia originating from gerbera leaves (p < 0.05). Germination tended to be higher at 20°C than at 25°C, although in general the decrease of germination in time tended to behighe r at 20°C than at 25°C (p < 0.1). After having been present on a cucumber plant for 31 days, 66 to 90% of conidiawer estil labl et ogerminate .Fo rconidi atransferre d from poinsettiaan dgerber aplants , germination levels were on average 62- 64 %an d 27- 45% , respectively, after 31days . Whiteflynymphal mortality Whitefly nymphal mortality differed significantly between host plant species (p < 0.001). Mortality oncucumbe r wasver yhig h andremaine dhig hdurin gth ecours eo f theexperiment , incontras t withmortalit yo ngerber ao rpoinsetti a (Fig.5.3) .Th epercentag e deadnymph swa s highesto ncucumbe rplants ,decreasin gonl yfro m about99 % (day0 )t o90 %o nda y3 1(20°C ) or 88%o n day2 9 (25°C).O ngerber a mortality decreased from 73% (day0 )t o 37%o nda y 31 (20°C) or decreased from 96% to 69% on day 29 (25°C). Mortality was lowest on poinsettia leaves,decreasin g from on average56 %(da y0 )t o4 %o n day3 1(20°C ) or to 17% (25°C).Th einitia lmortalit yo fT. vaporariorum onpoinsetti a waslo w (52- 59% ) compared toinitia l mortality oncucumbe r (99%) orgerber a (72- 95% ) (p< 0.001) .Th e decrease over timei n mortalityo f nymphs oncucumbe rleave si s significantly smaller than onpoinsetti a or gerbera. It seems that mortality at 25° C was higher than mortality at 20°C, although differences are small, but due to the lack of replicates atglasshous e scale, this could not be tested 74 Persistence ofAschersonia aleyrodis 100 ; • 80 | * o | 60 ; cucumber TTTTf-- £ - V ^9 CO T3 T c on L ° o o r o 60 r T—V-^_ ——1_ -- - - J _ c T o 40 "•4—• co c gerbera • 'E 20 ; 1— m =-^yV 0 » 1 100 ; 80 ' •------. : _<>;... s" I 60 i T poinsettia 40 j 20 \ j T ^-^F-f 0 ' 0 5 10 15 20 25 30 days after inoculation Figure 5.2:Mea n percentage germination ofA . aleyrodis conidia from 0 days until 31 days on leaf surface of cucumber, gerbera andpoinsetti a measured attw o temperatures;b ydirec t observation of agar imprint of leaf (= germination onlea f surface): - -V - - :20°C ,—•— : 25°Co robservatio n after 24hour sincubatio n on WA at 25°C( =germinatio ncapacity) :- - O - - : 20°C,—#— : 25°C.Vertica lline srepresen t standarderror so f means. Discussion The ability of an entomopathogen to persist in the habitat of its host is important for the effectiveness of naturally occurring and introduced pathogens (Jaques, 1983). Our results indicatetha tA. aleyrodisi sabl et osurviv e welli nth ehos thabita t andt okil l itshos t overa 75 Chapter 5 100 -e- ^—g- • —^- -•- rrfi 80 60 cucumber 40 20 Co" 0 o^ • ^ 100 - - - •• J- - - - • - - - -T_-_- -±- • ? 80 5 1 I -- "1~ - • E 60 "T" "" " lo • .e gerbera Q. 1 . T E 40 G >» c 20 +^ !c 3 0 100 80 60 ft 1 poinsettia 40 ^T-i- 20 _ T " -- -I ± s 5 "5" D- "•- -c-H- -o--*rr-J-—^ 5 10 15 20 25 30 days after inoculation Figure 5.3:Mea npercentag e whitefly mortalityb yAschersonia aleyrodis ( - - D- - :20°C , - -: 25°C) after presence of conidia on abaxial leaf surface of three whitefly host plants (cucumber, gerbera and poinsettia) in days between inoculation (day 0) and first possible contact. Vertical lines represent standard errors of means. period of at least one month. The persistence of conidia, determined in this study by germination capacity and nymphal mortality is, however, largely influenced by host plant species. Nymphal mortality and germination capacity of conidia seemed to be positively correlated, asha sbee n suggested byFargue set al. (1988) .I n our study oncucumbe r conidial germination capacityan dwhitefl y mortalityremaine d high for atleas t onemonth . On gerbera 76 Persistence ofAschersonia aleyrodis both declined, suggesting a less favourable environment for persistence of conidia. When poinsettia was used as a host plant for T. vaporariorum, nymphal mortality declined considerably in time, despite the high germination capacity of conidia. Since the initial mortality (day0 )o npoinsetti a was alsolow , otherfactor s than survival of conidia are likely to play a role. The results on poinsettia are not only limited to 71 vaporariorum, since comparable resultswer eobtaine dfo r silverleaf whitefly, B.argentifolii (S . vanVoors t &E . Meekes,unpubl .results) .Additionally , notonl ydi dconidi a of A. aleyrodisremai n viable for over amont h on poinsettia, but also conidia ofA . placenta showed the same ability and the same trend: a slow decline in viability on poinsettia and a steeper decline on gerbera (A. Tapaninen &E .Meekes ,unpubl .results) . Toasses sgerminatio ncapacit yo fconidi ao nth elea f surface,a naga rimprin ttechniqu e wasused ,rathe rtha n alea f washing technique.Th elatte rmetho dinvolve scountin go fcolon y forming units.Othe r microorganisms maycoloniz e the agarbefor e thepresenc e of the slow growing A. aleyrodisi s noted, soth e leaf washing technique is not very accurate (Fransen, 1987).I naddition , the agar imprint methodprovide s information about conidial 'behaviour' onth elea f surface. Iti sremarkabl e thatconidi adi dno tgerminat e inhig hnumber so nth elea f surface, althoughthe ywer eexpose dt oa nea rsaturate denvironmen tfo r4 8hours .O n artificial substrates likeparaffi n waxo rcellophane ,hig hpercentage s of conidia dogerminat e at93. 9- 100%an dfre e waterwa sno tnecessar y (Fransen, 1995),an dwhe n conidia arespraye d upon whitefly nymphs,the ygerminat ereadil y(persona lobservation ,E .Meekes) .I ti sno tye tknow n whether low germination onlea f surface isdu et o lack of stimulation or due toinhibitio nb y thepresenc eo f secondarychemical so nth elea f surface and/orphysica l factors. Incontras tt o A. aleyrodis conidia, up to 56%o f those ofV. lecanii germinated oncucumbe r leaves when incubated for 24hr sa thig hhumidit y (>95%) (Verhaar, 1998).I ngeneral ,A. aleyrodisseem s to 'sit-and-wait' rather than germinate and grow towards itshost . For this fungus, which is specialized on Aleyrodidae (Evans &Hywel-Jones , 1990),i t could be advantageous not to germinatei nabsenc eo fit shost ,especiall ysinc egermling s aremor e sensitivet o unfavourable conditionstha nnon-germinatin g conidia(Sussman , 1968).I naddition ,conidi ao fAschersonia spp.ar emainl ydisperse d bywater , soi f theygerminat e under thesecircumstances , theyca n loosethei r capacity to infect their host. The same fungal strain differed in its persistence on the three plant species, which could be due to anumbe r of reasons.First , the composition of thephyllospher e microbiota differ amongan dwithi nhos tplan t species.Som emicroorganism s are exclusively associated withon eplant ,other s are more opportunistic (Sinha, 1971).Whe n leaves age,difference s in microflora occur: thenumbe r andcompositio n of thephylloplan ebiot a increases toreac ha maximuma tsenescenc e(Cook e& Rayner , 1984).Severa linteraction sexis tbetwee n microbes 77 Chapter 5 onth elea f surface, sucha scompetitio n for nutrients,antibiosi s and mycoparasitism (Elad et al.,1996) ,an dth elea fmicrobe sma youtcompet eth eentomopathogens .Informatio n aboutth e microorganisms occurring on cucumber, gerbera and poinsettia is limited to fungal plant pathogens andintroduce d microbes.Cucumber , andt oa lesse rexten tgerbera , are susceptible toa grea tnumbe ro ffolia r diseases (Smith etal, 1988;Asselber g etal., 1996) .I nthi srespect , they differ from poinsettia, of which only one foliar disease is known, powdery mildew (Oidium sp.) (Motte & Unger, 1995; Koike & Saenz, 1998), indicating that either the microorganisms aresuccessfull y outcompetingth epathogen sor ,mor elikely ,tha tth ephysica l and/or chemical characteristics of the leaf create an environment unsuitable for growth of many fungi. Secondly,th echemica lcompositio n ofth ehos tplan t speciesca n differ substantially. Chemical substrates leached from the leaf mayac t as anutrien t source for fungi, while leaf cuticlescontai ncomponent stha tca nals ob eexploite d (Cooke& Rayner , 1984).O nth e other hand,leave so fcertai nplant sma yexud efungistati c substances,whic hca ninhibi t germination orrestric t germ-tube growth.Cuticula r waxesca ncaus e these effects also (Blakeman, 1971). Leaves of poinsettia contain cytotoxic triterpenoids (Smith-Kielland et al., 1996) and leaf extracts havebee n used as asee d treatment against nematodes and fungi (Khan etal., 1996 ; Sebastian &Gupta , 1996).Difference s in chemical plant substances mayals ohav e an effect on the host insect, thus influencing fungal infection indirectly. The insect diet can influence infection of theColorad opotat obeetl e byB. bassiana,renderin g the insect more susceptible (Hare& Andreadis , 1983).Funga linhibitor sals oproduce db yth eplan tma yprotec t theinsec t (Ramoska & Todd, 1985). Legaspi et al. (1996) showed that lacewinged flies feeding on silverleaf whitefly reared onpoinsetti adi d notreac h thepupa l stage,whic h was suggested to be due to accumulation of detrimental plant compounds inth e whitefly nymphs. Thirdly, the morphology of the host plant and its canopy characteristics may play a role. Leaf surface features, such as size and shape, surface topography and canopy characteristics varyamon gplan t species.Cucumbe rha slarge ,hair yleaves ,wherea spoinsetti a has smaller, rather smooth leaves.Th e cucumber canopyi s open ascompare d with gerbera, which is arosett e typeplant . These features affect the leaf boundary layer, in which insect- fungus interactions takeplace . Broad and/or hairy leaves have athicke r boundary layertha n narrowleave san dleave swit ha smoot hsurface . Butals otemperature ,placemen to fleave so n aplan t(sunn yo rshady )an dturbulenc ewil linfluenc e thethicknes so fthi slayer .Th ehumidit y in this layer is a gradient controlled by the stomata and the thickness of this layer, since exchangeo f gasesdepend s ondiffusio n processes (Burrage, 1971;Ferr o& Southwick , 1984). Themicroenvironmen to f thefungus , andespeciall yrelativ e humidity, is of great importance toit slongevity .Fo rman yfung i theintermediat ehumiditie s (50-70% RH)ar esuppose d tob e 78 Persistence ofAschersonia aleyrodis moredamagin g to their survival than both extremes, for conidia aswel l asgermling s (Clerk &Madelin , 1965;Sussman , 1968;Diem , 1971). In contrast to the results of Fransen (1995), temperature did not greatly affect the germination capacity and nymphal mortality by A. aleyrodis.Franse n (1995) showed that germination capacity of conidia on cucumber leaf at 25° Cdecrease d from 90%o n daynin e to 20% on day 15 p.i. and this was thought to be due to increased microbiota on the leaf surface causedb yearl y senescence.I nou r experiments, senescence ofpoinsetti a and gerbera leaves hardly occurred during the experiments. In cucumber, senescence of leaves was observed andth eleave s agedearlie r at2 5° Ctha n at2 0°C .Thi sdi dno toccu rbefor e 35day s p.i., which could explain the differences in germination capacity between Fransen's (1995) experiments and ours.I n our experiment a slightlyhighe r germination capacity and whitefly mortality wasobserve d at2 5 °Ccompare d with2 0 °C. A. aleyrodis isbein g evaluated for the control of whitefly in glasshouses, where the environment is less extreme than in open field situations. In its natural environment, (sub)tropical forest ecosystems, where bright sunshine alternates with clouds and rain, A. aleyrodisha st oovercom eextrem ecircumstances .Conidi ao fA .aleyrodis ar ebrigh torang e duet opigments ,mainl ycarotenoid s (Eijk etal., 1979) .Resistanc e of sporest osola rradiatio n is often related to presence of these pigments, which may enable pigmented spores to withstand shortperiod so fexposur et oradiatio nbette rtha nhyalin espores ,althoug h variation among and within species exists (Fargues etal., 1996). Aschersoniaaleyrodis i sno t known to produce resting spores or other structures to overcome unfavourable circumstances, such asthos eproduce d bysevera l entomophthoralean species (Uziel &Kenneth , 1991; Gindin& Ben-Ze'ev, 1994),bu t other strategies mayb e involved. Verticillium lecaniii s able to infect aphids on glasshouse chrysanthemums up to 14day sp.i. ,wit h aconidia l half-life of 4 days (Gardner etal., 1984) ,bu ti tsurvive ssubstantiall y longero nsporulatin g aphidcadaver s(Hall , 1981). Inconclusion , thepersistenc e ofA .aleyrodis conidiao ncucumbe r and gerbera leaves over longperiod s of time is apositiv e aspect of thisbiocontro l agent. Since newly emerged whitefly adults move upwards from the old canopy to younger leaves for feeding and oviposition (van Lenteren & Noldus, 1990), reapplication of the entomopathogen to new canopywil lstil lb eneeded .I twil lb einterestin gt okno wi fothe rentomopathogeni c fungi are able to survive in similar situations and will useth e same strategy to do so.Futur e research maysho wi fothe raspect si nth efunga l infection process,a swel la spersistence ,ar e influenced byhos tplan t species. 79 Chapter 5 Acknowledgements We would like to thank A.Tapanine n (HAS,Delft , NL) for her initial work on this subject, J.Tolsm a(PBG ,Aalsmeer ,NL )fo r hisvaluabl eassistanc e andd eBoe r(Centr efo r Biometry, WUR, NL) for advice on statistical analysis.W e are verygratefu l toE.A. M Beerling (PBG, Aalsmeer, NL), G.J. Bollen (Lab. for Phytopathology, WU, NL) and the Entomology PhD groupfo r criticallyreviewin gearlie r versionso f themanuscript .Thi sresearc h was financially supported byth eDutc hProduc t Boardfo r Horticulture (ProductschapTuinbouw) . 80 6 Roleo fphyllospher e climate on different host plants in the interaction between entomopathogenic fungi and the two pest species Trialeurodes vaporariorum andAphis gossypii ' Abstract Canphyllospher ehumidit yexplai n differences ininsec t mortality duet o entomopathogenic fungi on different crops? This was tested for cucumber, gerbera, tomato and poinsettia on whichmycosi so fgreenhous ewhitefl y (Trialeurodes vaporariorum) andcotto n aphid(Aphis gossypii) were determined in relation to host-plant climate. Phyllosphere humidity was estimated using climate and host-plant parameters. Hydrophobicity of the leaves and crop density were alsotake n into account. On cucumber, gerbera and tomato, the fungi Aschersoniaaleyrodis an dA. placenta caused over90 %mortalit y inwhitefly , while Verticillium lecaniicause d 50%mortality . On poinsettia, whitefly mortality was significantly lower for all three fungi (ca. 20%). The percentage of aphids infected by V. lecanii or M. anisopliae was significantly higher on gerbera than oncucumber . WithV. lecanii the highest percentage mycosis was obtained. The cucumber phyllosphere was more humid than the one on the other crops. This cannot explain higher aphid mycosis on gerbera compared to cucumber, norlowe r whitefly mortalityo npoinsettia .Th efac t thatpoinsetti a leaves weremor ehydrophobi c than the other leavesma yoffe r anexplanatio nfo rth eobserve dlowe rwhitefl y mortality, butchemica l host- plant aspects may also play a role. Our results underline the importance of the first trophic level (plant)fo r entomopathogenic fungi in integratedpes t management programs. 1 Submitted asMeekes ,E.T.M ./ Beerling,E.A.M. , Joosten,N.N. ,an d Fransen,J.J . Role ofphyllospher eclimat eo ndifferen t hostplant si nth einteractio nbetwee n entomopathogenic fungi and the twopes t species Trialeurodes vaporariorum andAphis gossypii. Chapter 6 Introduction The efficacy of entomopathogenic fungi asmicrobia l control agents may vary with the crop species on which these mycopesticides are applied (e.g.Hall , 1985;va n der Schaaf et al., 1991;Bolckman set al., 1995 ;Meeke s etal., 2000).Thi sma yb ea consequenc e of chemistry, physical structure and/or growth characteristics of thehos t plants and their canopies.Whe n testing mycopesticides in glasshouses, it is often assumed that the variation in efficacy on different cropsi sprimaril y causedb ycrop-dependen t phyllospherehumidit y (van der Schaaf et al, 1991; Bolckmans et al., 1995). The phyllosphere environment of the fungus, and especially therelativ ehumidity , iso f paramount importance toit s effectivity and persistence (Diem, 1971; Drummond et al., 1987). Ambient relative humidity, temperature and air turbulence willinfluenc e theboundar ylaye ri nwhic h theinsect-fungu s interaction takesplac e (Burrage, 1971;Ferr o& Southwick , 1984).However ,phyllospher ehumidit yan dthicknes so f the leaf boundary layer are also affected by crop-specific features such as size, shape and position ofleaves ,lea f surface topography (hairs,waxes ,stomata , veins),lea f areainde x and photosyntheticactivit y(transpiration ) (Ferro& Southwick , 1984).Physica lplan tfeature s may alsoinfluenc e insectbehaviou r(Dixon , 1987;Klingauf , 1987)whic hma yaffec t theinteractio n betweeninsec t andfungu s (Furlong& Pell , 1996).Furthermore ,plan t chemistryma y affect, positively or negatively, the insect-fungus interaction (Elliot etal., 2000). Wedetermine dinsec tmortalit ydu et oentomopathogeni cfung i onfou rglasshous ecro p species:cucumber , gerbera, tomatoan dpoinsettia . Thefung i Aschersoniaaleyrodis Webber , A. placentaBerkele y &Broom e (deuteromycetes: coelomycetes) and a whitefly-pathogenic strain of Verticillium lecanii (Zimm.) Viegas (deuteromycetes: hyphomycetes) were tested against the greenhouse whitefly, Trialeurodes vaporariorum Westwood (Homoptera: Aleyrodidae) on these four crops. On cucumber and gerbera, an aphid-pathogenic strain of V. lecanii and Metarhizium anisopliae (Metschnikoff) Sorokin (deuteromycotina: hyphomycetes) were used against cotton aphid, Aphis gossypii Glover (Homoptera: Aphididae). Both insect species are extremely polyphagous and amongst the principal pests in arang e of crops (Byrne etal., 1990a ;Blackma n &Eastop , 1984).Greenhous e whiteflies andcotto n aphidsdiffe r inthei rrelationshi p withthei r hostplant , asa resul t of differences in behaviour andlif e history.Whitefl y nymphs aresedentary , exceptfo r thefirs t stage,whil e all aphid stages aremobile .I ngeneral ,th emor esuitabl eth ehos tplant ,th eles smobil e theinsec t will be, which is valid for both insect species (van de Merendonk & van Lenteren, 1978; Knudsen etal., 1994).Anothe r difference istha tth egeneratio n time ofviviparou s aphidsi s usually shorter than that of whiteflies. On cucumber for instance the generation time of greenhouse whitefly at2 0° Cwa sca .3 6day s(Meekes ,unpubl .res.) ,wherea sfo r cotton aphid at 20 °Cth e developmental time on cucumber was 5t o 7day s (van Steenis, 1995). Efficacy 82 Role of phyllosphere climate ininsec t mycosis ofentomopathogeni cfung i islikel y tob e influenced differently bythes e contrasting insect - host plant relationships. WithA. aleyrodisan dA. placenta, fungi specialisedo nAleyrodidae ,goo dresult shav e been obtained in controlling whiteflies (Fransen, 1987; Meekes et al., 1996), as with a whitefly-pathogenic strain ofV. lecanii(va n der Schaaf etal., 1991). An aphid-pathogenic strain ofV. lecanii has been extensively tested asa contro l agent of aphids, including cotton aphids,wit hgoo dresult sdependin go nambien trelativ ehumidit y(e.g. Hall& Papierok , 1982; Helyer et al., 1992).M. anisopliaei s able to infect many aphid species (Butt et al., 1994; Milner, 1997). Here we tested the hypothesis that phyllosphere humidity can explain differences in insect mortality duet o insect-pathogenic fungi onfou r different crop species.W e measured different climatean dhost-plan tparameter s offou r different cropspecies ,i norde rt o calculate the vapour pressure deficit (VPD) in their phyllosphere. By measuring humidity and temperature inth eproximit y ofth elea f andth elea f temperature itself, thephyllospher e VPD canb eestimate d (e.g. Goudriaan &va n Laar, 1994).Direc t measurement of VPDwithi n the leaf boundary layeri simpossible ,a sth emeasuremen t itself would disturb this layer (Ferro& Southwick, 1984).Simultaneousl y with the climate measurements, greenhouse whitefly and cotton aphidmortalit ydu et oentomopathogeni c fungi wasdetermine d onth ehos t plants,an d correlated toVPD .I naddition ,hydrophobicit y ofth eleave s andcro pdensit ywer edetermine d and related to VPD. By studying more than one insect-fungus interaction in the same experiment,w eaime da texceedin gth elimit so fon einsec to rfungu s species,t oobtai na bette r understanding of therol eo f phyllosphere climate in their interaction. Materials and Methods Hostplants The host plants used were the ornamentals gerbera (Gerbera jamesonii cv. Serena) and poinsettia (Euphorbiapulcherrima cv. Goldfinger) and the vegetables cucumber(Cucumis sativus cv. Profito) and tomato (Lycopersiconesculentum cv . Moneydor). The plants were individually grown in pots and placed together on ebb and flow tables to form a crop. For poinsettiathi sresulte d in8. 3plants/m 2,fo r theothe rcro pspecie si n7. 2plants/m 2.Th egerber a plantswer eflowerin g duringth eexperiment ,whil eother splant swer evegetative .Fo rpractica l reasons, the two vegetables were lopped twice above the fourth leaf and lateral shoots were pinched out. This resulted in plants with a maximum height of one metre, in contrast to growers' common practice, where plants can grow to several metres. All plants were insecticide- and fungicide-free. 83 Chapter 6 Insects Trialeurodes vaporariorum, the greenhouse whitefly, was maintained on gerbera plants (Gerberajamesoniihybrids) .Aphis gossypii, thecotto naphid ,originate dfro m tomatoan dwa s rearedo ncucumbe ran dgerber aplant sfo rman ygeneration s(ca .2 months) .Al linsec tculture s werekep t in screened cages at2 1± 2 °C in glasshouses. Forth eexperiment , 50whitefl y adultso fbot h sexes wereconfine d in clipcage s (0 2 cm) on young leaves. The females were given the opportunity to oviposit for 24 hours and werethe nremoved , whichresulte di n5 0( ± 1.8 se)firs t andsecon dinsta rnymph safte r 16 and 18day sa t2 0 °C(o ntomat oan dgerber arespectively) .Th ewhitefl y populations from gerbera wereno tpreconditione d toth eexperimenta l hostplants ,sinc epre-conditionin g of greenhouse whitefly toa les ssuitabl ehos tplan t usually doesno t alterth e suitabilityo f thatspecifi c host plant (Thomas, 1993).Cotton-aphi d infested leaves werepicke d four and three weeks prior to theexperiment . Cucumber-reared aphids werepu t on experimental cucumber plants,an d gerbera-reared aphids wereuse d onexperimenta l gerbera plants,5-1 0 aphidspe r plant. Fungi Threefunga l strains,Aschersonia aleyrodis KV107, A. placenta Apl and Verticillium lecanii KV01,wer eapplie d against whitefly.Aschersonia aleyrodis originate d from Aleyrodidae in Colombia (W. Gams, The Netherlands) and A. placenta was isolated from Dialeurodes cardamomi(Homoptera :Aleyrodidae )fro mIndi a(S .Selvakumaran ,India )respectively .Thes e two Aschersonia strains were passaged through whitefly hosts twice. Sporulating colonies from pureculture so npotat odextros eaga r(PDA ,Difco ) wereuse dt oinoculat emille tculture s for massproductio n ofconidi a(Fransen , 1987).Culture swer eincubate d at2 5° Can dL16:D 8 (artificial light),an dconidi awer eharveste d from 2-3wee kol dcultures .Culture so fV. lecanii KV01originate dfro m thecommercia lproduc tMycota l(Kopper tBV ,th eNetherlands) ,whic h contained 1010conidia/gra mformulate d product.Th eproduc t wassuspende d indemineralise d waterwhic h wasplate d onPD Aan dincubate d at2 3° Cfo r 10days .Conidi ao f allthre e fungi weresuspende d indemineralise d watercontainin g0.05 %v/ vTwee n 80(Merck )a sa spreade r (Fransen, unpubl. res.) and suspensions were diluted to 107conidia/ml . On average conidial germination exceeded 90%o n water agar (Agar-agar, Merck). Aphids were treated with two fungal species: V. lecanii KV71 and Metarhizium anisopliae 299981. Cultures ofV. lecaniioriginate d from the commercial product Vertalec (Koppert BV, the Netherlands) and contained 109 blastospores/gram formulated product. M.anisopliae wa sisolate dfro m Homoptera (Trinidad) (A.K.Charnley ,UK) .Bot hfung i were passaged through the cotton aphid twice, and the fungi were subsequently cultured onPD A andincubate d at2 3°C .Conidi a were harvested from one totw o week old plates and diluted 84 Roleo f phyllosphere climate ininsec t mycosis to 1 x 107 conidia/ml with 0.03% Tween 80 (Butt et al., 1992). On average conidial germination exceeded 80%o nwate r agar. Experimental design Theexperimen t wascarrie d outi nfou r 150m 2glasshous e compartments, each containing 16 ebban dflo w tables (4.0m x 1.8 m) ascommonl y used for potted plants.Eac h compartment inthi ssplit-plo t design experiment was divided into two sections (A andB )wit h twoblock s of four (A) and two (B) tables. Crops were randomly assigned to tables within a block. In section "A", an experiment was carried out with greenhouse whitefly onfou r pairs of tables (=fou r plant species),eac h with 52plants ,excep t for poinsettia where therewer e 60plants . In section "B", an experiment was carried out with cotton aphids on two pairs of tables (cucumber and gerbera) again with 52 plants. All tables were divided into four plots for different treatments, which were randomly allocated. These plots consisted of four experimental plants,surrounde d bybuffe r plants of the same species. Treatmentprocedure Eachexperimenta lplan twa sspraye donc ewit ha D eVilbis shan dspraye r(Va nde rKui pLtd. , Utrecht,th eNetherlands ) using compressed air (ca. 3x 103 hPa);run-of f of suspensions was avoided. For whitefly-infested plants, one whitefly-infested leaf per plant was sprayed with one of the following applications: a) 1x 107conidia/m lA . aleyrodis,b ) 1x 107conidia/m l A.placenta, c) 1 x 107conidia/m lV. lecanii KV01, or d)0.05 %Twee n (application control), respectively.Poinsetti a formed almostn one w leavesdurin gth eexperiment .Hence , whitefly- infested leaves were found in the top of the poinsettia plant, whereas in the other crops, whitefly-infested leaveswer efoun d inth elowe rpart s (seeals ounde r 'phyllosphereclimate') . Whitefly mortalityo nth etreate dlea f wasassesse d after twoweeks ,b ywhic h time surviving insects had become pupae (one leaf per plant, four plants per plot).Number s of desiccated nymphsan dmycose dnymph s (orangenymph sb yA .aleyrodis orA .placenta o ropaqu ewhit e nymphsb yV. lecanii)wer edetermined .Becaus eo fpossibl eoverla pbetwee nnatura lmortalit y and mycosis,th eoveral l mortality wasuse d for analysis. Aphid-infested plantswer espraye dentirel yeithe rwith :a ) 1 x 107conidia/m lV. lecanii KV71, b) 1x 107conidia/m lM. anisopliae,o r c) 0.03%Twee n (application control),o r d) were left unsprayed. After oneweek , oneuppe r and onelowe r leaf perplan t (see alsounde r 'phyllosphere climate') werecollecte d and storedi nplasti cbag s at4 °C.A t this temperature, no new cases of mycosis will develop and aphid reproduction will be negligible, making it possible to examine leaves at a later date. Unfortunately, many aphids died under these 85 Chapter 6 circumstances andtherefor e onlyth epercentag e aphids withvisibl e signso f mycosiscoul db e used for analysis. Ambientclimate Theclimat ewithi n thefou r glasshouses wasmonitore d using aspiration boxes (Fluxon) with an electronic humidity sensor and a PT100 element for air temperature measurement. Temperature and humidity levels were recorded every 5 minutes during the course of the experiment andexpresse d astwo-hou r averages.Th eeb ban dflo w tables wereequippe d with a heating system (common practice for potted plants (Vogelezang, 1993),whic h was set at 20 °C,givin ga naverag etemperatur e of 19.6° C(rang e 18.9- 20. 9°C) .Th erelativ ehumidit y (RH)range d from aminimu m of 54.8% toa maximu mo f 73.1% .Experiment s were carried outunde r aregim e of 10hour s light and 14 hours dark. Phyllosphere climate In November and December, relative humidity (RH) and temperature were measured atth e first fully grown leaf andon eo f the oldest green leaves,usin g ahandhel d thermohygroscope (Rotronic Hygroskop DV-2) containing anelectroni c humidity sensor and aFI T0 0elemen t for air temperature measurement. The sensor was fixed in a stand, which was adjusted toa distance of 1c m from the underside of a leaf. On consecutive days the climate in gerbera, poinsettia, tomato and cucumber was determined, the latter was used as the reference. Measurements werecarrie dou to nth ereferenc e crop and one of the othercrop s onon etabl e per crop in all four glasshouse compartments and on twoplot s per table.Thi s was repeated three times on different days. Leaf temperatures were measured using an infra-red camera (Heimann,S925) . Theai rVP Dwa sestimate d from themeasure d airR Han dtemperatur e {e.g. Goudriaan &va n Laar, 1994).Th ephyllospher e VPD wasestimate d from leaf temperature, leaf widthan dai rVP D(Goudriaa n &va nLaar , 1994;Goudriaa n pers.comm.) .Formula san d assumptions are given in theappendix . Plantmorphology As a measure for crop architecture and density, the decrease in photosynthetically active radiation (PAR) was determined (Sunfleck Ceptometer CEP40,Decago n Devices Inc.,US) . The larger the difference in PAR above (set at 100%) and below the canopy, the denser the crop. The PAR above and below the canopies of gerbera, poinsettia, tomato and cucumber were determined; the latter was used asreference . This was repeated three times during the experiment for each comparison (two replicates per glasshouse compartment, four compartments perday) . 86 Role ofphyUospher e climate ininsec t mycosis Thehydrophobicit yo f leavesfro m thefou r plant specieswa sassesse d qualitativelyb y placing 10u l of water or 0.05%v/ v Tween (spreader) in demineralised water onth e abaxial sideo f aleaf .Leave swer eo f similar aget o those used in theexperiment . Immediately after placing a droplet, photographs were taken. The shape of water droplets on a leaf surface reflects the hydrophobicity of the leaves (Holloway, 1970). The rounder the droplet, the smallerth econtac tbetwee nth edrople tan dth elea f surface, andth emor ehydrophobi cth elea f surface. Byaddin g a spreader, thecontac t between droplet and the leaf surface will become larger. Statisticalanalysis Fractionswhitefl y mortality (four leavespe rplot ) andfraction s aphidmycosi s (twouppe ran d twolowe r leaves perplot ) were arc-sin square-root transformed. Transformed mortality and mycosis were analysed using a two-way ANOVA with (fungal) treatment and host-plant species as independent variables, and glasshouse compartments, blocks, tables and plots as block treatments (Genstat 5 release 3.22, Payne etal., 1987). Subsequent differences were comparedwit ht-tests ,i nwhic hprobabilit y(a )wa sadjuste d toth enumbe ro ft-test s performed (n)( 87 Chapter6 Tween V.lecani i A.aleyrodi s A.placent a d d d d d d 100 2^80 1 60 o E sj* 40 £ § 20 Gerbera Cucumber Poinsettia Tomato hostplan t Figure 6.1: Whitefly mortality on cucumber, gerbera, poinsettia and tomato caused by application control (0.05% Tween), V. lecanii KV01,A. aleyrodis and A. placenta. Bars followed by the same letter are not significantly different (comparisons can only be made within the same treatment or host plant species;fo r a = 0.001, lsd for arcsin\/(fraction) transformed data = 0.145. Results Greenhouse whitefly Whitefly mortality varied according to host plant (Fig. 6.1; adjusted percentages, back- transformed from the arc-sin square root transformed fractions). All fungi caused a significantly lower mortality on poinsettia (< 21%, and not significantly different from the control) than the other three crop plants.Whitefl y mortality in the control was significantly higher on poinsettia than on other host plants. Both Aschersonia species performed significantly better than V.lecanii o n cucumber, gerbera and tomato. All fungal treatments caused a significantly higher mortality than the control treatments on all crops, except for poinsettia. Cotton aphid Aphid mycosis on upper and lower leaves of gerbera and cucumber was generally low, but significant host-plant differences did exist (Fig.6.2 , adjusted percentages, back-transformed from the arc-sin square root transformed fractions). On the upper leaves, both crop and treatment had significant effects on aphid mycosis; no significant interaction between these 88 Roleo f phyllosphere climate ininsec t mycosis V7, untreated Tween M. anisopliae V. lecanii Upperleave s Lowerleave s 15 W w o o 10 E Q. CO a a a a a a a jzi Gerbera Cucumber Gerbera Cucumber hostplan t Figure 6.2: Cotton aphid mycosis on upper and lower leaves of cucumber and gerbera caused by control (untreated), application control (0.03% Tween),M. anisopliae and V.lecanii KV71. Upper leaves, main effect 'crop': cucumber (A) and gerbera (B); for a = 0.025, lsd for arcsin ^(fraction) transformed data = 0.123. main effect 'treatment': control (a),applicatio n control (a),M. anisopliae (ab) and V.lecanii KV71 (b);fo r a = 0.004, lsd for arcsin v'(fraction) transformed data = 0.141. Lower leaves, interaction between 'crop' and 'treatment': gerbera/control (a),gerbera/applicatio n control (a),gerbera/M . anisopliae (a),gerbera/V . lecanii (b), cucumber/control (a), cucumber/application control (a), cucumber/A/, anisopliae (a),cucumber/V . lecanii (a); for a = 0.003, lsd for arcsin\/(fraction) transformed data = 0.173 effects was found. In general, the percentage of mycosed aphids on upper leaves of gerbera was significantly higher than onuppe r leaves of cucumber. Furthermore, V.lecanii induce d significantly higher mycosis than both controltreatments , whereasM. anisopliaeresulte di n anintermediat e percentageo fmycose d aphids.O nth elowe r leaves,w edi dfin d a significant interaction between cropan dtreatment .A highe rpercentag eo f mycosed aphidswa sfoun d on gerbera plants treated with V. lecanii compared to the other treatments and compared to cucumberplants .Mycosi sfoun d inbot hcontro lgroup si sprobabl yth eresul to f contamination during the experiment andnatura l infection of thepopulation . Phyllosphere climate Thephyllospher e ofcucumbe rwa smor ehumi dtha n that ofgerber a andtomat o (Tab. 1a) .O f these crops, only the humidity of the cucumber phyllosphere (VPD = 1.48 -1.73 hPa, RH= 93.6 - 92.6% at 20 °C) exceeded 92%,th e value critical for germination of fungal conidia 89 Chapter6 (Walstad etal., 1970;Ferron , 1977;Fransen , 1987;Chandle r etal., 1994).Th e upper leaves of all three plants were more humid than the lower leaves due to heating of the tables (Vogelezang, 1993). A significant interaction between crop and leaf level was found when comparing phyllosphereVP Do fpoinsetti a andcucumbe r (Tab. lb).Lowe rleave so fpoinsetti a wereles s humid than upper leaves, andbot h were less humid than thecucumbe r leaves. Theuppe r leaveso fpoinsetti awer einfeste d withwhitefl y andwer euse dfo rassessin gwhitefl y mortality. Thehumidit yo fth epoinsetti a upper-leaf phyllosphere (VPD =1.7 6 hPa,R H= 92.7 %a t2 0 °C)exceede d 92%,th evalu ecritica lfo rgerminatio n offunga l conidia (Walstad etal., 1970; Ferron, 1977;Fransen , 1987;Chandle r etal., 1994) . Table 1:Estimate d vapour pressure deficit andmatchin g RH(a t2 0°C ) inth ephyllosper e atuppe r andlowe r leaf level; a:onl y main effects, that ishost-plan t species andlea f level,ar esignificant , b:interactio n between both main effects is significant. a Main effects Comparison Host-plant species VPD [hPa] RH[%] Leaf level VPD [hPa] RH[%] Cucumber - Cucumber: 1.73 a*1 92.6 Upper 1.85 A2 92.0 Gerbera Gerbera: 2.42 b 89.8 Lower 2.30 B 90.4 Cucumber - Cucumber: 1.48 0s 93.6 Upper 1.74 A" 92.4 Tomato Tomato: 2.41 b 89.8 Lower 2.16 B 90.8 b Upper leaf: Lower leaf: Comparison Host-plant species VPD [hPa] RH[%] VPD [hPa] RH[%] Cucumber - Cucumber: 1.10 c5 95.3 1.35 a 94.4 Poinsettia Poinsettia: 1.76 b 92.7 2.46 c 90.2 *numbers followed byth esam e letters areno tsignificantl y different; typographically different letters indicate different setso fcomparisons ; oc'sar ecorrecte d forth enumbe r of comparisons made resulting inth e following lsd's: '0.117,20.117,30.143, "0.143 and50.240. Relationshipbetween phyllosphere climate andmortality/mycosis Themortalit yo rmycosi sdat afo reac hplan t- insec t- fungu s combination wererelate d toth e calculated phyllosphere VPDs (Fig.6.3) .Belo w 1.9 hPa almost no germination of fungal conidia will take place (Walstad etah, 1970;Ferron , 1977;Fransen , 1987;Chandle r etal., 1994), andtherefor e wedivide d VPD in low,rangin g from 0-1.9 hPa(9 2- 100%R H at 20 °C),an d high, when larger than 1.9hP a (vertical linei nFig . 6.3).Fift y percent mortality wasuse d asa roug hdistinctio n between high and low mortality andmycosi s (horizontal line inFig .6.3) .I fphyllospher e climateexplain sdifference s ininsec t mortality and mycosis,the n 90 Roleo f phyllosphere climate ininsec t mycosis low VPD relates to high mortality and high VPD to low mortality, and data points will be withinth equadrant s H0.Fo r moretha nhal f ofth edat a setsthi si sno t thecas e(Fig .6.3) . 100 Hr 8$ 10 4 11 o o 80 E Q. 60 as A12 •oc CO H £ 40 r 9A to o E 4 45 £• 20 >17 ,19 CD A6 13 15 J8 J4 16 .20 1.00 1.50 2.00 2.50 3.00 VPD (hPa) Figure 63: Relationship between phyllosphere vapour pressure deficit (VPD) and whitefly mortality or aphid mycosis. Numbers indicate different plant - insect - fungus combination (see table below). H0 = mortality or mycosis > 50% and VPD < 1.9 hPa, or mortality or mycosis < 50% and VPD > 1.9 hPa. Insect Fungal Cucumber leaf Poinsettia leaf Tomato leaf Gerbera leaf species species higher lower higher lower higher lower higher lower whitefly • A. placenta 1 4 7 10 A. aleyrodis 2 5 8 11 V.lecanii KV01 3 6 9 12 aphid • V. lecaniiKV7 1 13 15 17 19 M. anisopliae 14 16 18 20 Plantmorphology In this experiment more light was transmitted through the cucumber crop than through the gerbera andpoinsetti a crops, meaning thatth e cucumber cropwa sles s dense (Fig. 6.4). No difference in lighttransmissio n between cucumber and tomatocrop s was found. Thehydrophobicit y of leavesfro m thefou r different plant species was studied byth e shapeo f waterdroplet splace d onth elea f surface (Fig. 6.5). The roundest droplet was found 91 Chapter6 on poinsettia, indicating that it has the most hydrophobic leaves of all four host plants. No largedifference s existedi nshap eo fth edroplet sbetwee n tomato,gerber aan dcucumber .Als o the effect of a spreader on the shape of water droplets was studied (results not shown). All dropletswit h spreader were lessroun dtha n theircounterpart s without spreader, but the same differences existed in shape of thedroplet s between the cropspecies . Gerbera Cucumber Poinsettia Cucumber Tomato Cucumber 5 10 15 20 25 30 Light transmission through canopy [%] Figure 6.4: Percentage transmitted light (photosynthetically active radiation, PAR) through the canopy: cucumber (dark grey) compared to gerbera, poinsettia, and tomato. Vertical lines represent standard errors of mean. Discussion Whitefly mortalityan daphi dmycosi sclearl yvarie d according to host-plant species (Fig.6. 1 and 2).W eals ofoun d significant differences inphyllospher e VPD between thosehos t plants (Tab. 1).Th e question is whether these differences in phyllosphere humidity between host plantsca nexplai n differences in mortality and mycosis. If this wouldb eth ecase ,on e would expect a high mortality or mycosis to be associated with to a high humidity, and a low mortality or mycosiswit h alo w humidity (H„i n Fig.6.3) . Low levelso f whitefly mortality were found onpoinsetti a (4,5,6i nFig . 6.3), versus high mortality levels ontomat o (7,8,9i n Fig. 6.3) and gerbera (10,11,12i n Fig.6.3) , while onpoinsetti a amor ehumi d phyllosphere climate was found than on theothe r crops.Fo r aphids,bu t not for whitefly, themycosi s on gerbera(17,18,19,2 0i nFig .6.3 )wa shighe rcompare dwit hmycosi so ncucumbe r(13,14,15,1 6 92 Roleo f phyllosphere climate ininsec t mycosis C>j-._'/..•' •.'i^ 1 Figure6.5 : Shape of water droplets onlea f surface of A: poinsettia, B: cucumber, C: tomato, and D: gerbera, indicating inhydrophobicit y of theseleaves .Ba r = 0.34 mm. in Fig. 6.3), whereas cucumber had a morehumi d phyllosphere climate than gerbera. From theseresult sw econclud etha tVP Ddifference s betweenhos tplant scanno texplai n differences inmortalit y andmycosis .Phyllospher eclimat ei n amor e open crop iscommonl y believed to beles shumi dtha ntha ti na dens ecro p(va nde r Schaaf etal, 1991;Bolckman set al., 1995) . Inou rbetween host-plantexperiment s thelatte rargumen t does not hold, sincecucumbe r had amor ehumi d phyllosphere climate despite amor eope n crop structure, whilepoinsetti a and gerbera,wit h amor edens ecro p structure,ha d alowe rhumi dphyllospher e climate (Fig.6.4) . Thecounter-intuitiv e resulttha tVP Ddifference s betweenhos tplant scanno t explain differences inmortalit yan dmycosis ,i sdiscusse dbelow .Physica l andchemica laspect so fhos t plantsma yclarif y ourresults . Physicalaspects of the host plant Aphyllospher eVP Do f 2.4hP a (~ 90%R H at2 0 °C) ongerber a andtomat o was apparently sufficient forbot hAschersonia speciest ocaus ea hig h whitefly mortality (7,8 , 10, 11i nFig . 6.3),comparabl et omortalit ylevel s oncucumbe r (VPD< 1.9 hPa orR H> 92% ; 1,2 i nFig . 6.3). Although many entomopathogenic fungi need aVP D of ca.1.9hP a orlower , that isa n RHo fmor etha n 92%(a t2 0°C) ,optima l germination andgrowt htak eplac e at0 hP a(100 % RH;(Gillespi e &Crawford , 1986;Milne r &Lutton , 1986;Fransen , 1987).Th ehig h whitefly 93 Chapter 6 mortality ongerber aan dtomat osuggest stha tth ecalculate d VPDi s anoverestimatio n of the realphyllospher e VPD.On e shouldrealis e that the VPD data shown here are an averageo fa whole leaf and thus obscuring differences within alea f caused by roughness (e.g.hairiness , venation),stomata ,an dothe rlea f characteristics (Holloway, 1970).Furthermore ,th e influence of the insect on the phyllosphere climate and the microclimate of the insect itself (Kramer, 1980;Ramoska , 1984)whic h could affect fungal germination and growth, areno ttake n into account. Vidal etal. (1998) also did not find differences in whitefly (Bemisia argentifolii) mortality oncucumbe r and tomatoeithe r when testingPaecilomyces fumosoroseus. Further, whitefly mortality on cabbage was not different from cucumber and tomato (Vidal et al., 1998).I n thoseexperiments , however, thehighl yfavourabl e greenhouse climate, which has apositiv eeffec t onphyllospher ehumidit y(Ferr oet al., 1979 ;Gaede , 1992),ma yhav eblurre d differences between crops. For poinsettia, the calculated phyllosphere VPD may have been underestimated (= phyllosphere humidity overestimated). On this crop, all three fungi (4, 5, 6 in Fig. 6.3) caused whitefly mortality nohighe r than 21%, whereasbot hAschersonia specie s were able to cause high whitefly mortalities on cucumber, gerbera and tomato under similar or less favourable phyllosphereVPDs . (Bailie etal., 1994a; 1994b) found that the transpiration rate for poinsettiai sver ylo wcompare d with other ornamentals, andlea f stomatal resistance was considerablyhighe rfo r poinsettia (highertha nr s= 100s/m , aswa sth eassumptio n inth eVP D formula, seeAppendix) .Thi sresult s in ahighe r VPD than wecalculated , which mayinhibi t fungal growtho rma yeve nb edetrimenta lt oth efungus . Inthi srespec t thehydrophobicit y of theleaf ,whic hi sinfluence d byit swa xlaye ran dhairiness ,i sals olikel yt opla ya considerabl e rolei nth eactua lphyllospher ehumidity .O fal lplan tspecie suse dhere ,poinsetti aha dth emos t hydrophobic leaf surface. Water droplets on this leaf stayed compact, even when a spreader wasused ,i ncontras t to droplets on gerbera, cucumber ortomat oleave s (seeFig .6.5) . Apart from negativelyinfluencin g thedistributio n ofconidi ao nth elea f surface ofpoinsettia , it will also negatively affect the formation of athi n film of fluid on this leaf surface after spraying orcondensation .Thi s couldi n turndecreas e thelikelihoo d that conidia getint o contact with the insect host and couldnegativel y affect germination of conidia. Less virulent fungi will never give rise to high mortality regardless of the humidity conditions,an dthi sma yhav ebee nth ecas efo rM. anisopliaeagains taphid so ncucumbe r(1 4 in Fig. 6.3), V.lecanii agains t aphids (13i nFig . 3)an dV. lecaniiagains t whitefly (3i nFig . 6.3). Differences in virulence can explain the difference in whitefly mortality between the Aschersonia species andV. lecanii(va n derPa set al., 1996;Franse n unpubl.res.) . Entomopathogenic fungi differ considerably in their humidity requirements for germination. Some fungi may need a saturated environment, like V. lecanii (Hall, 1981). 94 Role of phyllosphere climate in insect mycosis Others,suc ha sM. anisopliae (Walstad etal., 1970 )an dA .aleyrodis (Fransen, 1995),ar eabl e togerminat eunde rslighd ydrye rconditions .Whe nhumidit ycondition s areunfavourable , the fungus remains inside the insect (Ferron, 1985). Since aphid mycosis was detected using externalcues ,th e actualmycosi s wasprobabl yunderestimated . Asa consequence , onewoul d expectt oobserv e ahighe r aphidmycosi s oncucumbe r than on gerbera because ofth ehighe r phyllosphere humidity of cucumber. However, theobserve d mycosis ongerber a was higher thano ncucumber ,whic hsuggest stha ta differenc e inphyllospher ehumidit ywa sno tth emai n sourceo fdifferenc e inaphi dmycosi sfoun d betweenthes etw ocrops .Th elo wobserve d aphid mycosiscause db yM. anisopliae compared totha tcause db yV. lecaniima yb eexplaine d by different humidityrequirement s ofthes efung i forprotrusio n outsideth einsect ,alon gwit hth e slower killing and sporulation of M. anisopliae compared with V. lecanii (Hall, 1980). In general, successful control of aphidpopulation s isknow n to depend largely onth e epizootic potential of thefungu s (Hall &Papierok , 1982;Hall , 1985) and thus on sporulation, which mayexplai n thelo w aphid mycosis in this short-term experiment. Chemical aspectsof the host plant Chemicalaspect so fhos tplant s areknow n topla y arol ei n insect-fungus interaction aswel l as physical aspects (Hare &Andreadis , 1983;Cost a & Gaugler, 1989;Brow n et ah, 1995; Hajek et ah, 1995).Plan t species differ substantially in their chemical composition, among whichth enatur ean dquantit yo f toxicsubstance s inthei rphloe m sap(Harborn e etal., 1999) . For instance, some tomato cultivars produce tomatine which was found to be toxic to Beauveria bassiana (Costa & Gaugler, 1989). Also, leaves of poinsettia contain cytotoxic triterpenoids (Smith-Kiellandet al., 1996 )an dextract so fthes eleave swer euse d against fungi and nematodes (Khan etal., 1996; Sebastian &Gupta , 1996). However, the viability ofA . aleyrodisconidi awa sno t affected by aprolonge d stay onpoinsetti a orcucumbe r leaves, so no direct effect of toxins on conidia was apparent (Meekes et al., 2000). Persistence of V. lecaniiconidi ao ngerber aleave si sconsistentl yhighe rtha no ncucumbe r(Beerling ,unpubl . res.), which may explain the lower performance of V. lecanii against aphids on cucumber comparedwit hgerbera . Sinceinfectio n of aphidsdepend slargel yo npic ku po fconidi a (Hall &Papierok , 1982),phyllospher e chemicals maypla y alarge r role inpersistenc e of fungi, in contrast to infection of aprincipall y sessileinsec t like whitefly. Also, the influence of host plant via the insect is important (for overview, see Berenbaum, 1988).Fruits , roots or seeds of the Cucurbitaceae contain cucurbitacins which werefoun d toprotec tth espotte dcucumbe rbeetl eagains tM. anisopliae(Tallam yet al., 1998), butw euse d ales sbitte rcucumbe r cultivarwhic h containsver ylo w amounts of cucurbitacins (A.Janssen , pers.comm.) .Whiteflie s (B.argentifolii) reared onpoinsetti a areunsuitabl e as 95 Chapter6 food for lacewings, whichha sbee n attributed tonutritiona l inadequacy ort o accumulation of detrimental plant compounds in the whitefly (Legaspi et al., 1996). A similar interaction betweenfirs t and third trophic level mayexis ti n entomopathogenic fungi on this host plant, hence explaining thelo w whitefly mortality on poinsettia. Developmental rate, reproduction and mortality of whitefly and aphids are also influenced byhost-plan t species (Minks &Harrewijn , 1987;va n Lenteren &Noldus , 1990). Thecontro lmortalit yo fgreenhous e whitefly wassignificantl y highero npoinsetti arelativ et o cucumber, gerbera or tomato (see also van Lenteren &Noldus , 1990). A less suitable host plant, as reflected by higher natural mortality, could result in an increased susceptibility to entomopathogens,du et ostres so fth ehos tinsec t(Steinhaus , 1958).I ncontrast ,thi scoul dals o result in an decreased susceptibility to entomopathogens compared to insects from a more suitable host plant, since these 'healthier' insects better support growth of natural enemies (Schultz & Keating, 1991; Hoover et al., 1998). However, here low fungal infection is probablyno tlinke d tohost-plan t quality, sinceanothe rwhitefl y (B.argentifolii) exhibited the samelo winfectio n levelso npoinsetti aa sgreenhous ewhitefl y (E.Meekes ,unpubl .res.) ,whil e poinsettia is a very suitable host plant for B. argentifolii (van Lenteren & Noldus, 1990; Fransen, 1994). Thedevelopmenta lan dreproductio nrat eo fcotto naphi dwa shighe ro ncucumbe rtha n on gerbera (Beerling, pers. obs.). Aphids were more likely to escape fungal infection on cucumberdu et oa shorte rexposur e time(Clanc y& Price , 1987),whic hinvolve s alsoth elos s of inoculum by moulting. In addition, a higher reproduction had consequences for the percentage survivingaphids ,especiall ywhe n onerealise s thateve n infected aphids givebirt h (Hall, 1976).Furthermore , amor efavourabl e host plant doesno tencourag e mobility of the insect andtherefor e picku po fconidi a(Hal l& Papierok , 1982).Thes e aspectscombine d may explain the higher aphid mycosis on gerbera than on cucumber, in spite of the higher phyllosphere humidityo n cucumber. Concluding remarks Poinsettia init svegetativ ephas ei sattacke d byfe w pest species andonl yon efolia r pathogen: powdery mildew (Oidiumsp. ) (Motte & Unger, 1995; Koike & Saenz, 1998). Cucumber, gerbera andtomat ocarr yman ymor epest s anddisease s(Anonymous , 1999).Th elea f surface of poinsettia seems therefore amor e hostile environment for pests and plantpathogens ,an d maybe also for entomopathogenic fungi. From an evolutionary perspective, the differences found inefficac y offung i ondifferen t plantspecie ssuppor tth etheor ytha tsom eplan t species may have invested in creating an environment which fosters entomopathogens, thereby obtainingbodyguard s toprotec t them againstherbivore s asa nindirec t defence. Thiscontrast s 96 Roleo fphyllospher e climate ininsec t mycosis withothers ,lik epoinsettia , which mayhav e invested rather in defending themselves directly against plant pathogens at the cost of potential bodyguards (Price et al., 1980; Dicke & Sabelis, 1988;Sabeli set al, 1999; Elliot etal, 2000). Ourresult s demonstrate thatth eefficac y of entomopathogenic fungi doesno t depend only on the interaction between the entomopathogenic fungus and the target insect. Of overriding importance is the tritrophic interaction with the host-plant species, as shown by variationi n successo fentomopathogeni c fungi ingreenhous epes t controli n different crops. Differences in phyllosphere humidity between the crops cannot explain our results. Direct chemical effect ofth ehos tplan t onfung i andindirec t chemical effect viath e insect probably accountlargel yfo rth eobserve ddifference s inefficac y ofentomopathogeni cfung i on different hostplants .Thes efinding s areo fsignificanc e for(developmen tof ) efficient microbialcontrol . Asi sth ecas ewit hpredator s andparasitoid s (Barbosaet al, 1982;Sabeli s &Jong , 1988;Ve t &Dicke , 1992),ou rresult sunderlin e theimportanc e ofconsiderin gth efirs t trophicleve l (the plant) for entomopathogenic fungi in integrated pest management programmes. Acknowledgements We thank J. Tolsma, H. Schiittler and J. van Zoest for their technical assistance, and W. de Boer(Centr efo r Biometry,NL )an dJ.Th.W.H .Lamer sfo r their adviceo nstatistica l analysis. A.K. Charnley (University of Bath,UK )kindl yprovide dM. anisopliae299981 , andR . van der Pas (Koppert Biological Systems, NL) and S. Selvakumaran (Indian Cardamom Res. Institute,India )kindl yprovide dAschersonia cultures .Specia lthank sw eow et oM.W .Sabelis , S. Elliot and A. Janssen (Population Biology, UvA, NL), J.C. van Lenteren and the Entomology PhD group (Lab. for Entomology, WU, NL), J. Goudriaan (Plant Production Systems Group,WU , NL) and A.F.G. Jacobs (Meteorology and Air Quality, WU, NL) for valuablediscussion s andcriticall yreviewin g earlier versions of this paper. Thisresearc h was funded by the Dutch Product Board for Horticulture (Productschap Tuinbouw) and the European Union (AIRCT94-1352) . 97 Chapter6 Appendix Thevapou rpressur edefici t in thephyllosphere , VPDleaf isbase d on Goudriaan and vanLaa r (1994) and Goudriaan (pers.comm.) . VPDleaf= VPD air -r bvyR n / pCp +(s+yr bvl rbh )AT [hPa] where': - vapourpressur e deficit in air, VPDair - (1- RH air/100) es(Tair) [hPa] saturation vapour pressure,e s(mr) =6.107 exp(17.4 Tair/(239.0 +T air)) [hPa] - leafboundar y layerresistanc e toheat ,r bh =100i/(w/v) [s/m] leaf width", w =0.23 (cucumber) , 0.11(gerbera) , 0.09(poinsettia) , 0.04(tomato ) [m] - local windspeed, v = 0.1(i nglasshous e onlyconvection ; Gates (1980)) [m/s] -psychrometri ccoefficient , y =0.67 (at 20 °C) [hPa/°C] 3 - volumetric heat capacity of air, pCp =1200 [J/m °C] - slopeo f thesaturatio n vapourpressur ecurve ,s =1.45 (at 20 °C) [hPa/°C] -resistanc e toexchang e of water vapour'", rbv = 2 (0.93r bh) [s/m] - temperature difference, AT =T kaf- T air [°C ] - netradiatio n absorbed perlea f area,R„ =AE + H [W/m2] 2 sensible heat losspe r leaf area,H =AT pCp / rbh [W/m ] latent heat loss, AE= (sH + 3)/ y* [W/m2] 2 dryingpower , 6 = VPDairp Cp / rbh [hPaW/m ° q r*=V(r, +rJ/r bh [hPa/°C] total leaf resistance, r, =r s rc/(rs +rj [s/m] -*• r,- rs (cuticular resistance,r cnegligible ) [s/m] leaf stomatal resistance, rs = 100 (when stomata are open) [s/m] Bold: measured parameters. Leaf width was measured on leaves of similar age as those used in the experiment. The resistance to water vapour (rbv)i s twice asbi g asth e resistance tohea t (rbh), since water vapour can only diffuse from the underside of the leaf (stomata), whereas heat can diffuse from both sides of the leaf 98 7 Relativehumidit y and host-plant species influence mycosiso fwhitefly , Trialeurodes vaporariorum, byAschersonia aleyrodis, A. placentaan d Verticillium lecanii1 Abstract Theinfluenc e ofrelativ e humidity (RH),host-plan t species andthei r interaction on mycosis of greenhouse whitefly, Trialeurodes vaporariorum, byentomopathogeni cfung iAschersonia aleyrodis, A. placenta and Verticillium lecanii was studied. Experiments, conducted on poinsettia, gerbera and cucumber at50 %an d 80%RH ,showe d aclea r host-plant effect. On cucumber and gerbera all fungi caused significantly higher mortality compared with the control, whereas whitefly mortality on poinsettia remained low and was not significantly different from thecontrol .A highe rambien tR Hresulte d ina highe rmortalit ycause db yeithe r of the three fungi. BothAschersonia spp. performed significantly better than V. lecanii.I n subsequent experiments an additional period of 0, 3, 6, 12, 24 or 48 hours of high RH was applied after treatment with A. aleyrodis and V. lecanii. Mortality caused by the fungus increased with a longer additional period of high RH in combination with a higher ambient humidity.However , thehost-plan teffec t exceeded theeffec t of ambiento radditiona lhig hR H period; whitefly mortality washighes t on cucumber, intermediate on gerbera and lowest on poinsettia.O npoinsetti abot hfung i onlyha d a significant effect onwhitefl y mortality after a 48hr shig h RHperiod . Ongerber a andcucumbe r A. aleyrodisperforme d better under dryer conditionstha nV. lecanii. Relations between ambient RH,funga l species and the host-plant environment are discussed. 1 Submitted as Meekes, E.T.M, Joosten, N.N., Fransen, J.J. and Lenteren, J.C. van. Relative humidity and host-plant species influence mycosis of whitefly, Trialeurodes vaporariorum, byAschersonia aleyrodis, A. placenta andVerticillium Chapter7 Introduction Variousabioti cenvironmenta lcondition sinfluenc e fungal pathogen- hos tinsec tinteractions . Thefungu s ishighl ydependen t on suitableenvironmenta l conditions during different phases init slif e cycle,viz. theinitia lpar to fth einfectio n cycle,spor egerminatio n on theinsec t host andpenetratio n of itscuticle , and, later on,durin g protrusion of thefungu s from the insect's body andconidiogenesi s (Ferron, 1977). Very high RH levels (> 95%) are required for spore germination and sporulation outsideth ehos tinsec t(Hallswort h& Magan , 1999).However ,entomopathogeni c fungi differ considerablyi nthei rhumidit yrequirement sfo r germination. Somefung i maynee da saturate d environment or free water, like Verticillium lecanii or Hirsutella thompsonii (Hall, 1981; McCoy, 1981).Others ,suc ha sBeauveria bassiana, Metarhizium anisopliae(Walsta d etal., 1970)o rAschersonia aleyrodis (Fransen, 1995), cangerminat ea tlowe rR Hlevels .Moreover , ambientR Hi sles scritica l thanpresume d previously, sinceth eR Hi n the micro-environment surrounding the spore influences germination more than ambient RH (Marcandier & Khachatourians, 1987;Ravensber g etal., 1990) . Fungal processes take placei n theboundar y layer surrounding the leaf (orinsect ) in which the airi sundisturbed . TheR H atth elea f surface isdependen t onth e diffusion ratea t which water vapour is transferred through the boundary layer. The humidity in the phyllosphere istherefor e determined bystomata laperture ,th ethicknes so fth eboundar y layer and their interaction (Burrage, 1971). The aperture of the stomata is mainly influenced by environmental factors, sucha slight ,C0 2 concentration, water status of theplant , differences inambien thumidit yan dlea f temperature (Jones, 1986).Th e thickness of theboundar y layer islargel ydetermine d byhost-plan t characteristics,suc ha slea f sizean dshape ,hairines so fth e leaf and placement of the leaf on the plant, but also byturbulenc e and ambient temperature (Ferro &Southwick , 1984).Thu s ambient andphyllospher e RHar e correlated. Inth etritrophi csystem ,hos tplan t- whitefl y -entomopathogeni c fungus, thehos tplan t can influence the efficacy of the fungus not only via the micro-environment, but also by production of allelochemicals or a combination of both. Allelochemicals can stimulate or inhibit the fungus directly (Blakeman, 1971)o reffect s mayoccu r via theinsec t (Gallardoet al, 1990;Haje k etal, 1995). Previous research showed varying results on poinsettia, gerbera and cucumber (previous chapters;Meeke s etal., 2000) ,whic h mayb e explained by tritrophic interactions. Thefunga l speciesstudie di nthi spaper ,A. aleyrodis,A. placenta andV. lecanii, differ inthei r host spectrum andhumidit yrequirements . Species of the genusAschersonia ar erestricte d to whiteflies and scaleinsect s (Evans &Hywel-Jones , 1990),wherea s V.lecanii i s a common soilinhabitin g fungus, which is able to survive saprophytically and infect abroade r rangeo f 100 humidity andhost-plan t influence onwhitefl y mycosis insects,lik eaphids ,thrip s and whiteflies (Hall, 1981).I ti s alsoknow n asa mycoparasit e on powdery mildew (Verhaar etal., 1998)o rrus t (Zouba &Kahn , 1992).I nit snatura l habitat, (sub)tropical forest ecosystems, A. aleyrodis only causes epizootics under very humid conditions (Wolcott, 1955), since it is dependent on rain for splash dispersal (Samson & Rombach, 1985).However , when applied as amycoinsecticid e against greenhouse whitefly oncucumber ,A. aleyrodis wasabl et ocontro lth ewhitefl y population ata ambien tR Ho f50 % (Fransen, 1987b).Whe nV. lecaniiwa sapplie d againstwhiteflie s oncucumbe r ortomato , the eventual success seemed not tob ecorrelate d with ambient RH (Ravensberg etal., 1990). Inthi spape r wedescrib eth einfluenc e of ambientR Ho nmycosi so f whitefly on the host plants cucumber, gerbera and poinsettia at continues high or low RH, and at levels fluctuating arounda hig ho rlo waverag eRH ,wit hadditiona lperiod s ofhig hR Hafte r fungal treatment. Materials andMethod s Plantmaterial The effect of host plant on efficacy of entomopathogenic fungi was studied on three plant species: cucumber (Cucumissativus cv. Profito) gerbera {Gerberajamesonii hybrids) and poinsettia (Euphorbiapulcherrima cv. Goldfinger). Allplant s were free of insecticides and fungicides. The two youngest, almost unfolded leaves on each poinsettia, gerbera and cucumber plant were infested with greenhouse whitefly (seebelow) .Afte r infestation with whitefly, only thecucumbe r plants weretoppe d twice above the third leaf and lateral shoots werepinche d out. Whitefly host Trialeurodesvaporariorum wasmaintaine d ongerber aplant si nscreene dcage sa t21° Cunde r greenhouse conditions.T o obtain second/third instar whitefly nymphs,5 0 adults,bot h male and female, wereconfine d to clip-cages onleaves .Th e adults were given theopportunit y to oviposit for 24 hrs,whic h resulted in 188± 6 (SE),9 9 ±4 and 137± 6nymph s per leaf for cucumber, gerbera andpoinsettia , respectively. The average developmental period from egg tosecond/thir dinsta rnymp hvarie dbetwee n 18day so ncucumbe ran d2 2day so ngerbera .Th e relative contribution of the different nymphal instars toth epopulatio n was approximately 1 first: 6 second : 10thir d instar whitefly nymphs for all plant species at the time of fungal application. 101 Chapter7 Fungi Threefunga l strains,Aschersonia aleyrodis KV107,A. placenta Apl and Verticillium lecanii KV01, wereapplie d against whitefly.Aschersonia aleyrodis originated from Aleyrodidae in Colombia (W. Gams, The Netherlands) and A. placenta was isolated from Dialeurodes cardamomi (Homoptera: Aleyrodidae) from India (S. Selvakumaran, India) respectively (chapter 2).Thes e twoAschersonia isolate s were passed through greenhouse whitefly three timesbefor e augmentation. Sporulatingcolonie sfro m potato dextrose agar(PDA )wer euse d to inoculate millet-cultures for mass production of conidia. The fungi were cultured on autoclaved millet in Erlenmeyer flasks, which were closed with cotton, to allow aeration. Cultures were incubated at 25 °C and L16:D8 (artificial light). Conidia from 2-3 week old cultureswer eharveste db yrinsin gculture swit hsteril edemineralise dwate rcontainin g0.05 % (v/v) Tween 80 (Merck) as a spreader (Fransen, unpubl.). Suspensions were diluted to 107 conidia/ml. Cultures of V. lecanii KV01 originated from the commercial product Mycotal® (Koppert BV,th eNetherlands) ,whic h contained 1010conidia/gra mformulate d product. The product wassuspende d indemineralise d waterwhic h wasplate do nPD Aan dincubate d at2 3 °C for 10days . Conidia were suspended in demineralised water containing 0.05% Tween. Suspensions were filtered (Wattmann filter no. 1)an d diluted to 107conidia/ml . Treatmentprocedure Twom lo f theA. aleyrodis, A. placenta,V. lecanii or0.05 %Twee n suspensions werespraye d onth eunderside s of theleave susin g aPotte r spraytowe r (Burkard Manufacturing, UK).Th e 0.05%Twee n acted ascontrol .Leave s stayed attached toth eplan t until final assessment for mortality. After evaporation ofth ewater ,plant swer emove dt oth eclimat ebo x or glasshouse (seebelow) . Assessment of mortality took place after two (exp. 1)o r three (exps.2 and 3)weeks . Mortality wasdivide dint omortalit ycause db yth efungu s and that byothe rcauses . Anymp h was considered infected by the fungus when it turned opaque orange (A. aleyrodis, A. placenta)o rwhit e(V. lecanii). Anymp h wasconsidere d 'deadb yunknow ncauses' ,whe nth e insectha ddesiccate d and no apparent sign of fungal infection was visible. Overall mortality was used for analysis. Experiment1: Continuous highor lowRH Aschersonia aleyrodis,A. placenta, V. lecanii or Tween suspensions were applied to the leaves.Afte r spraying,plant s wereplace d ina climat ebo x (Fitoclima 801, Snijders Scientific, NL) at a high (80 ± 4%) or low (50 ± 6%) RH level and a temperature of 20°C (± 1°C) . 102 humidity andhost-plan t influence on whitefly mycosis Temperature andR Hwer emeasure d continuously.Pe rtreatmen t combination, twoleave spe r plant and five plants were used; plants were left uncovered. The experiment was repeated twice.Fina l assessment of whitefly mortality tookplac e two weeks after fungal application. Experiment2 and3: Variableperiod ofhigh RH Aschersonia aleyrodis, V. lecanii or Tween suspensions were applied to the leaves. After spraying,plant swer ereturne dt oth eglasshous e andwer eeithe rcovere d withplasti cbag s for 3,6,12,24o r4 8hr st ocreat ea conditio no f9 5-100 % RH orlef t uncovered( 0hr shig hRH) . Pertreatmen tcombination ,tw oleave spe rplan tan dsi xplant swer eused .Afte r removalo fth e plasticbags ,plant swer ekep to ntable swit h aneb b- flo w irrigation systemi n airconditioned glasshouse compartments (30 m2) set at 20°C. The temperature and RH in the crops were measured every 5minute s and averaged overtw ohou rperiods . The experiment was carried out twice: once in spring and once during the summer, bothtime si ntw oglasshous ecompartments .Durin gth esprin gexperimen t (exp.2 )th eaverag e ambient RH was 45.1% (± 7.1%SD ; min.: 31.7%, max.: 60.6%) and during the summer experiment (exp. 3)th e average ambient humidity was 88.7%( ± 8.9 %;min. : 80.1%, max.: 96.0%).Th etemperatur e waso naverag e20. 0± 0. 2 °C(SD )durin gth esprin gexperimen tan d 20.3 ± 1.0 °C during the summer experiment. Final assessment of whitefly mortality took place threeweek s after fungal application. Statisticalanalysis Fractions whitefly mortality per plant were arc-sin square-root transformed. Transformed mortalitydat awer e analysed using athree-wa y ANOVA with (fungal) treatment, host-plant species andambien tR H (high/low- exp . 1,short/lon g- exp .2 o r 3) asindependen t variables andclimat ebo xo rglasshous ecompartmen t asbloc ktreatmen t (Genstat5 releas e3.22 ,Payn e et al., 1987). Subsequent differences were compared with t-tests, in which probability was adjusted toth enumbe r of t-tests performed (0.05/n). Results Experiment1: Continuous highor low RH Ingeneral , athig hambien tR H whitefly mortalitycause db yth efung i washighe rtha n atlo w ambientRH ,bu tno ti nth econtro l (Tab.7.1 ;Fig .7.1) .A. aleyrodisan dA. placenta caused a higher mortality thanV. lecaniia t 50%a s well asa t 80%RH . A clear host plant effect on whitefly mortality was found. On poinsettia, all fungi causeda significantl y lowerwhitefl y mortality( s 25%mortalit yan dno tsignificantl y different 103 Chapter 7 VA Control • V.lecani i• A.aleyrodi s • A.placent a 100 Poinsettia 80 60 40 20 J.I 100 Gerbera 80 CO 60 40 j^ti 100 Cucumber 80 60 40 20 50 80 50 80 50 80 50 80 Relative humidity (%) Figure 7.1:Averag e overall mortality of T. vaporariorum in control (0.05% Tween), or caused by V. lecanii, A. aleyrodis or A. placenta on poinsettia, gerbera and cucumber at 50% and 80% ambient RH. Error bars represent standard deviation. 104 humidity andhost-plan tinfluenc e on whitefly mycosis from control) than on cucumber and gerbera (Tab. 7.1, Fig. 7.1). The control mortality however, wassignificantl y highero npoinsetti a thano nth eothe rhos tplants . A. aleyrodis and A. placentaperforme d significantly bettertha nV. lecanii,o ncucumbe ran dgerbera .Al l fungal treatments caused a significantly higher mortality than the control treatments on all crops, except for poinsettia. Table 7.1: The average whitefly mortality (%*) per treatment and ambient RH level or per treatment and host plant species (p < 0.001). Main-effects Treatments RH level Control V. lecanii A. aleyrodis A. placenta 50% 6.1 £**! 32.2 b 49.6 c MA c 80% 7.0 a 51.7 c 75.6 d 72.8 d Host plant poinsettia 19.7 Jj**2 25.3 b 24 b 22.5 b gerbera 1.9 a 52.5 c 79.1 d 80.7 d cucumber 3.3 a 49.2 c 86.4 d 80.0 d Adjusted percentages, back-transformed from the arc-sin square root transformed fractions; **: Numbers followed byth e same letter areno t significantly different, comparisons can onlyb emad e between treatments and within RH level or host-plant species; bold and normal letters indicate different comparisons; ': a = 0.0016, lsd = 0.116;2: a = 0.0008, lsd = 0.149; lsd's are only valid when data are arcsinv^fraction) transformed. Experiment2: Variableperiod ofhigh RH Thisexperimen tperforme d insprin g atlo w ambientR H(ca .45% ) showedtha tth elonge r the additional period of high RH, the higher the probability was for whitefly nymphs to dieo f infection. A host-plant effect was also evident here. On poinsettia, both V. lecanii and A. aleyrodis infected whitefly nymphs atlo w levels only.Mortalit y caused byV. lecaniiwa s not significantly different from thecontro ltreatment ,eve n after 48hr s additional high RH.I n contrast,mortalit ycause db yA .aleyrodis wa ssignificand ydifferen t fromth econtro ltreatmen t after 24an d4 8hr shig hR H(Fig .7.2) .O ngerbera , both fungi caused ahighe r mortality after 12 hrs high RH compared to the control mortality, A. aleyrodis causing a slightly higher mortality than V. lecanii (48 hrs high RH). On cucumber both fungi caused the highest mortality compared topoinsetti a andgerbera .Th ewhitefl y mortality caused byth e fungi was significantly highertha nth econtro lmortalit yi nal lcases ,eve nwithou t anextr aperio do fhig h RH,wit hth eexceptio n ofth eV. lecanii at 12hr sadditiona lhig hRH .Especiall yA .aleyrodis performed wellagains tT. vaporariorum andcause d 55% (0hr shig hRH )t o95 % (48hr shig h RH)mortalit y ata naverag e ambient RHo f 45% inth e glasshouse. 105 Chapter7 Control I V. lecanii A. aleyrodis 100 Poinsettia 80 60 40 20 c CD JHilF E 0 3 6 12 24 48 0 3 6 12 24 48 0 3 6 12 24 48 •c 100 CD Q. Gerbera X CD 80 O) C *i_ Q. 60 CO Z- 40 CO 20 •oc E mmMwMm ,l.j >. 0 3 6 12 24 48 0 3 6 12 24 48 0 3 6 12 24 48 »^- ••£ 100 Cucumber 80 60 40 20 0 3 6 12 24 48 0 3 6 12 24 48 0 3 6 12 24 48 Hourswit h highrelativ e humidity Figure 7.2: Average whitefly mortality in control (0.05% Tween) or caused by V.lecanii andA. aleyrodis on poinsettia, gerbera andcucumbe r atsi x different periods of highRH :0 , 3,6, 12, 24an d4 8hr si n spring (ambient RH45%) . Percentages areback-transforme d from the arc-sin square root transformed fractions. Bars followed byth e samelette rar eno t significantly different (comparisons can only be made within the same treatment, host plant species orR H level; a =0.0002 , lsd =0.264 ; lsdi sonl yvali d when data arearcsW(fraction ) transformed). 106 humidity andhost-plan t influence on whitefly mycosis Control V.lecani i IA .aleyrodi s 100 Poinsettia 80 60 40 20 c CD £MQEL E 0 3 6 12 24 48 0 3 6 12 24 48 0 3 6 12 24 48 CD 100 Q. X Gerbera CD i~ 80 o CD 73 _ O | T3 T3 T3 • E a E n HOI ° u 1 60 L ' o o o ^H^H^HI 40 ~ i CO ! •£ 20 O ; co co co co co co E i.jTjKzjmFzfcl,. 0 3 6 12 24 48 0 3 6 12 24 48 0 3 6 12 24 48 •= 100 Cucumber 80 60 40 20 -lU^V^r-^ 0 3 6 12 24 48 0 3 6 12 24 48 0 3 6 12 24 48 Hourswit hhig hrelativ e humidity Figure 7.3: Average whitefly mortality incontro l (0.05% Tween) orcause d by V.lecanii or A. aleyrodis on poinsettia, gerbera andcucumbe r at sixdifferen t periods of high RH: 0, 3,6 , 12,2 4an d4 8hr si n summer (ambient RH 85%).Percentage s areback-transforme d from the arc-sin square root transformed fractions. Bars followed by thesam e letter areno t significantly different (comparisons canonl y be made within the same treatment, host plant species or RH level; a = 0.0002, lsd = 0.232; lsd is only valid when data are arcsW(fraction) transformed). 107 Chapter7 Experiment3: Variableperiod ofhigh RH During theexperimen t performed insummer ,th e ambientR Hwa saroun d 85%.Als oher ew e found thatth elonge rth eadditiona lperio do fhig hRH ,th ehighe rth emortalit yfro m infection byV. lecanii andA. aleyrodis. Asi nth e springexperiment , ahost-plan t effect wasfoun d (Fig. 7.3). On poinsettia both fungi had hardly any effect on whitefly mortality, although now V. lecaniicause d asignifican t increasei nmortalit y after aperio d of4 8hr s high ambient RH, whereasA. aleyrodiscause da significantl y highermortalit y after 24an d4 8hr shig hRH .O n gerbera the twofung i performed equally well againstT. vaporariorum, this in contrast with the springexperiment . Oncucumbe rA. aleyrodisan dV. lecaniicause d thehighes t whitefly mortality compared to gerbera and poinsettia. A. aleyrodis, causing 81 to 97% mortality, performed againsignificantl y better thanV. lecanii,causin g 59t o90 %mortality . Discussion Hostplant -ambient humidity In this tritrophic system of host plant - whitefly - entomopathogenic fungus, the host-plant species appeared tob eo foverridin g importance in theeventua l success of microbial control. On poinsettia the period of high ambient RH (>95%) had to exceed 24 hrs to reach 50% mortality of T. vaporariorum, whereas on cucumber 50% to 70% mortality was reached without anadditiona lperio d ofhig hR Han deve nunde rrelativel ydr ycondition s (avg.45 % - 50% RH) (see Fig. 7.1 - 7.3). Rovesti et al. (1997) also found that only partial control of whitefly (Bemisiatabaci) o npoinsetti a wasachieved , althoughenvironmenta l conditionswer e very favourable for fungal growth. Vidal et al. (1998) found no host-plant effects when Paecilomycesfumosoroseus was applied against the silverleaf whitefly (B.argentifolii) o n cucumber, tomatoan dcabbage .However , since Vidal's experiments were carried out athig h ambient RH (near 100% ambient RH for 10 hrs) host-plant effects, if present, could be inconspicuous.O npoinsetti aP. fumosoroseus (Pakista n strain,courtes yo fL.A .Lacey )cause d similar low mortality levels as A. aleyrodis and V. lecanii (Fransen, unpubl.) and those mortality levels of bothT. vaporariorum andB. argentifoliiwer ecomparabl e to this study. The differences in mortality mentioned above could be related to differences in humiditya tth ehos tplant' s leaf surfaces, sincephyllospher ehumidit yi spartl ydetermine d by thehos tplan t itself and byit srespons e onenvironmenta l conditions (Burrage, 1971;Jones , 1986;Baili eet al, 1994a).Cucumber , withit slarger ,hair yleave s (vanLentere net al., 1995) , andgerbera , withit srosett etyp eo f plant,ma yhav e athicke r boundary layertha n poinsettia. Inaddition ,th einflu x ofhumidit yint oth eboundar ylaye ro fpoinsetti a mayb elow , sinceth e leaf stomatal resistance was considerably higher for poinsettia than for other ornamentals (factor 3 difference). Poinsettia also appeared to have a more hydrophobic leaf surface 108 humidity andhost-plan t influence on whitefly mycosis compared to cucumber and gerbera (chapter 6). All these differences in physical host-plant characteristics suggest that the phyllosphere humidity on poinsettia is lower than on gerbera orcucumber . Underhumi d conditions mycosed whitefly nymphso npoinsetti a only showed fungal protrusion from the insect's body at some humid places in the crop, in contrast with infected nymphso ngerber a orcucumbe r (Meekes,pers .obs.) .However , when phyllosphere humidity wascalculate d for thesehost-plan t species,th eobserve d differences did not explain thelo wwhitefl y mycosiso npoinsetti a compared tocucumbe ro rgerber a (chapter 6)an deve n under very suitable RH levels (89% RH, with additional 48 hrs 95 - 100% RH) whitefly mortalityo npoinsetti a didno treac hth e samelevel s aso ncucumbe r orgerber a (Fig.7.3) . So, although the above differences in plant characteristics will influence the effectivity of entomopathogenic fungi, theyar eprobabl yno tth esol eexplanatio n for thepoo r performance ofthes efung i on poinsettia. Hostplant -chemical aspects Chemical aspects of hostplant s areknow n topla y arol ei n insect-fungus interaction aswel l (Elliot etah, 2000;Poprawsk i etal, 2000).Plan tspecie sdiffe r substantially inthei r chemical composition, among which the nature and quantity of toxic substances in their phloem sap (Harborne etal, 1999).Poinsetti aleaves ,fo r instance,contai n cytotoxictriterpenoid s (Smith- Kielland etal, 1996)an dextract so fthes eleave swer euse dagains tfung i (Khan etal, 1996). However, since conidia of A. aleyrodisstaye d viable for over amont h on poinsettia leaves, direct toxic effects on conidia can likelyb eexclude d (Meekes etal, 2000;chapte r5) . Secondary plant substances may also affect the insect's fitness and through that its susceptibility topathogen s(Har e& Andreadis , 1983;Boucia set al, 1984;Ramosk a& Todd , 1985;Gallard oetal, 1990).Fo r instance,fruits , rootso r seedso fth eCucurbitacea e contain cucurbitacins whichwer efoun d toprotec t the spotted cucumberbeetl e againstM. anisopliae (Tallamy etal, 1998),bu t we used ales s bitter cucumber cultivar which contains very low amountso f cucurbitacins (A.Janssen ,pers .comm.) .Lacewing s (Chrysoperla sp.)feedin g on poinsettiareare dwhitefl y (B.argentifolii) didno treac hth epupa l stage,whic h was attributed to consumption and accumulation of detrimental plant compounds in the whitefly nymphs (Legaspi etal, 1996).A simila rinteractio n between first and third trophic level mayexis t in entomopathogenic fungi on this host plant, hence explaining the low whitefly mortality on poinsettia. The higher control mortality of greenhouse whitefly on poinsettia relative to cucumber orgerber a (Fig.7.1) ,i sno tlikel yt ohav eaffecte d theinfectio n levelshere ,becaus e another whitefly (B. argentifolii) exhibited the same low infection levels on poinsettia as greenhouse whitefly, while its control mortality onpoinsetti a is similar to thato n cucumber (Fig.9.3 ;chapte r9) . 109 Chapter7 Fungalspecies - ambient humidity Aswit hmos tothe rbiologica l control agents,entomopathogeni c fungi are limited bya n array of biotic and abiotic factors. Traditionally the ambient RH has been considered the most serious constraint on the use of fungi for the control of insects, as germination of conidia requires aR H above 90t o 95% (Gillespie &Crawford , 1986;Hallswort h &Magan , 1999). Ourresult s showedtha ta hig hambien tR Hpositivel yinfluence d mycosiso fT. vaporariorum. Iti sals ono tsurprisin gtha tshor tperiod so fnea rsaturate dhumidit yincrease d insectmycosis , since optimal germination and growth of many entomopathogenic fungi take place at near 100% RH (Schneider, 1954;Gillespi e &Crawford , 1986;Milne r &Lutton , 1986;Fransen , 1987a;Hallswort h &Magan , 1999). Aschersoniaaleyrodis andals oA. placentawer emor evirulen t oncucumber , gerbera andpoinsetti atha n V. lecanii, especiallyunde rrelativel y dryconditions . V.lecanii seeme d to be more sensitive to low ambient RH, as was found by Riba and Entcheva (1984). For V. lecaniith edail yperiod s ofhig hR Hha dt oexcee d 16hr st oenabl e thefungu s togro w and successfully infect greenhouse whitefly oncucumbe r andtomat o and atleas t 20hr s of high RH were needed to cause more than 50% mortality (Ekbom, 1981). However, V.lecanii isolateswer efoun d tovar yi nthei rhumidit yrequirement s (Drummondet al., 1987) .Exposur e to 12 hrs of high humidity on cucumber (Fig. 7.2) resulted in lower infection levels for V. lecaniitha n0 t o6 hr shig hhumidity .V. lecaniicolonize sth ehos tcuticl ebefor e penetration willtak eplac e (Schreiter etal., 1994;Askar y etal., 1999)an d mayb e more sensitive atthi s stage to the sudden dropfro m >95%t o45 % humidity byremova l of theplasti cbags . However, in spite of high humidity requirements for germination, fungal infection occurs ata widerang eo f ambient RHlevel s (Kramer, 1980;Ramoska , 1984;Marcandie r & Khachatourians, 1987, Fig 7.1-7.3), suggesting that the humidity at the surface of the host integument or on the leaf surface is sufficient for spore germination and host penetration (RavensbergefaZ., 1990;St.-Leger , 1991). Practical considerations Inglasshouse si nth eNetherland s wherepoinsetti a isgrow n during late summer and autumn, theambien tR Hi skep tbetwee n 70-85%,bu tambien thumidit yma yb ehighe rwhe nblack-ou t shields are used to obtain short days for initiation of flowering (Fransen, pers. comm.). However, even at an average ambient RHo f 89% (Exp. 3),whitefl y mortality byfung i was low (< 30%). In glasshouses with year round production of cucumber or gerbera, growers employa rang eo f climateregimes ,an dclimat econdition s varywit h season and from yeart o year(e.g. Kerssies, 1994;Di k& Elad , 1999).Ambien t humidityma yvar ybetwee n 35 - 95%, but at night ambient RHlevel s of 80-95% are easilyreached . This should be sufficient for 110 humidity andhost-plan tinfluenc e onwhitefl y mycosis whitefly controlwit hfungi ; maybewit hexceptio n ofperiod so ffros t ando nbrigh t sunnyday s when ambient RHca n dropt over ylo wlevel s (vanHoute n etal., 1995) .I n our experiments we applied the fungi only once, whereas for commercial products repeated applications are recommended (e.g. Wraight &Carruthers , 1999).Repetitiv e application would increase the mortalitylevel stha tw efound . With adiurna lfluctuatio n in ambient RHi n glasshouses,hig h duringth enigh t- lo w duringday ,repetitiv e application ofhig hambien tR Hperiod swoul d be anotheroptio n (Ekbom, 1981;Fransen , 1987a).However , artificially increasing the ambient RHi nglasshouses ,whic hha sbee n suggestedb ysom eauthor s(Helye r etal., 1992)woul db e unacceptable to most growers,becaus e of therisk t o stimulate plant-disease development. Ourresult sar ebase do nconidia l suspensionsamende dwit h0.05 %Twee na sspreader . Morerecen tdevelopment si nformulatio n ofconidi aha sle dt oth eus eo foi lcarriers . Oils are compatible with hydrophobic surfaces, such asinsec t and leaf cuticles, eliminating thenee d for wetting,stickin go rspreadin gagent s(Wraigh t& Bradley , 1996)an dthe yimprov e efficacy of, for instance, Metarhizium flavoviride under low moisture conditions (Bateman et al., 1993). Against the whitefly B. tabaci the use of oil formulations has been successful, increasing the efficacy of P.fumosoroseus on cucumber plants from no apparent infection (Tween formulation) to 80-100% infection (oil formulations) (Smith, 1994). Also oil formulations ofV. lecaniisho w promising results (Pas etal., 1998).I f a low RH at thelea f surface ison eo fth emai nreason swh ycontro lo fwhitefl y onpoinsetti a islow ,oi l formulation canb ea noption .I naddition ,oil sca nals ob einsecticida l (Helyer, 1993).I nfact , someoil sar e registered asinsecticide s (Wraight &Bradley , 1996).Possibl e sublethal effects of oils on the target insect could enhance the eventual effect of entomopathogenic fungi (Ferron, 1985; Quintella &McCoy , 1997). Our results clearly indicate that A. aleyrodisan d A.placenta are less dependent on ambientR Han dperfor m better against the greenhouse whitefly thanV. lecanii. A.aleyrodis andA. placenta willb e avaluabl e addition to biological control of whiteflies on crops like gerbera(chapte r 8)an dcucumber , especially with thelon gpersistenc e of thesefung i onlea f surfaces (Fransen, 1995;Meeke set ah, 2000;chapte r 9)an dth ecompatibilit y ofA . aleyrodis withth ewhitefl y parasitoid Encarsiaformosa (Fransen &va nLenteren , 1993;1994) . Ill Chapter 7 Acknowledgements We are grateful to J. Tolsma (PBG, NL) for the technical support and to R. van der Pas (KoppertBiologica l SystemsBV ,NL )an dS .Selvakumara n (IndianCardamo mRes .Institute , India)fo r kindlyprovidin gAschersonia cultures or infected insects for our research. We also thankE.A.M .Beerlin g (PBG,NL ) andth eEntomolog y Ph.D.grou p (WU,NL ) for critically reviewingearlie r versionso f thispaper .Thi sresearc h wasfinanciall y supported byth eDutc h Product Board for Horticulture (Productschap Tuinbouw). 112 8 Effect of gerbera cultivar onth e efficacy of entomopathogenic fungi against Trialeurodes vaporariorum^ Abstract Can host-plant characteristics influence the efficacy of entomopathogenic fungi? This was testedusin gtw ogerber acultivar swit hdifferen t plantarchitectur e andtrichom edensit yo nth e leaf:'Bianca 'havin gmor ean dlarge rleave s andfewe r trichomes/cm2 than 'Bourgogne'. Two entomopathogenic fungi, Aschersonia aleyrodisan d Verticillium lecanii,wer eteste dfo r the control of greenhouse whitefly, Trialeurodes vaporariorum. The fungi were applied in two concentrations, 106an d 107conidia/ml ,i n agerber a crop in aglasshous e situation. In'Bianca' ,bot hfung i causeda whitefl y mortalityu pt o 80%. Whitefly mortality was higherfo r 107tha nfo r 106conidia/ml .I n 'Bourgogne',V. lecaniicause d a significantly lower mortality than A aleyrodis.Althoug h no cultivar differences in whitefly development time were found, other characteristics, like natural mortality and build-up of the whitefly population, seemed to differ. Differences in mortality by the fungi in relation to whitefly characteristics and cultivar, including spray droplet distribution, are discussed. 1 Submitteda sMeekes ,E.T.M ,Joosten ,N.N. ,Fransen ,J J . andLenteren ,J.C .van . Effect of gerbera cultivar on the efficacy of entomopathogenic fungi against Trialeurodes vaporariorum. Chapter8 Introduction Biological control in floriculture is still difficult compared with biological control in fruit vegetables,du e to several reasons.First , morechemical s are available for ornamentals than for fruit vegetables, because of safety regulations for vegetables. In addition, due to thehig h number of cultivated ornamental species and cultivars, compared to vegetables, a high diversity in pest and disease problems, and avariet y of different environmental and cultural conditions exist. Secondly, in fruit vegetables only the fruits are harvested, which allows a higher population level of thepest , and some leaf injury can be tolerated. In contrast, most ornamentals arebein gsol dwit hflower s andleaves ,an dtherefor e shouldb efre e ofinjur y and presence of insects.Thirdly ,th e criterion of zerotoleranc e for both pest andbeneficial s has been used asa standar d for all ornamental products until now, though it isonl y needed asa n export requirement for afe w countries (Fransen, 1993;Gullin o &Wardlow , 1999). Currently,developmen to fIntegrate dPes tManagemen t(IPM )program sfo rornamenta l crops have a high priority, because of insecticide resistance problems and demand of the generalpublic .I nth eornamenta l cropgerber a (Gerberajamesonii), 80-85%o fth egrower s worksaccordin gt oIP Mguidelines ,whil ei twa sonl y 10%5 year sag o(M.J .va nde rMey ;F.R . van Noort, pers. comm.). In this crop only flowers are harvested and therefore cosmetic damage to leaves is acceptable within certain limits, and biological control ismor e feasible. Several parasitoids and predators are used commercially against greenhouse whitefly (Trialeurodes vaporariorum) in gerbera. However, for treatment of hot spots or when population levels exceed the action threshold, chemical control is imminent. Screening for natural enemies which are able to kill pest insects quickly, without affecting other natural enemies,i stherefor e animportan t lineo fresearch .Entomopathogeni c fungi canb ea valuabl e assett oexistin gbiologica lan dchemica lcontro lmeasure s(Lace yet al., 1996) .I nth elas t two decades several entomopathogenic fungi, such as Verticilliumlecanii, Paecilomyces spp., Beauveria bassiana, Metarhizium anisopliae and Aschersonia spp. (Fransen, 1987; Ravensberg etal., 1990 ;Malsa met al., 1998 ; Wraight etah, 1998), havebee nteste d for their potential to control T.vaporariorum, wit h good results. Their ability to infect insects with sucking mouthparts bypenetratio n of the cuticle, which other entomopathogens cannot do, make them suitablepathogen s of whitefly species (Fransen, 1990). Inth eNetherlands ,gerber ai sa neconomicall yimportan tcro p( >200,00 0ha )o fwhic h over60 0cultivar sar egrow ncommerciall y (F.R.va nNoort ,pers .comm.) .Currently , cultivars are mainly selected for colour and shape of their flowers and their post-harvest quality. However, they differ largely in other characteristics such as plant chemistry and plant morphology, likelea f shape, leaf size and trichomedensit y (Sutterlin &va nLenteren , 1996; Krips et al., 1999; Krips, 2000). These characteristics can influence the efficacy of natural 114 Cultivareffect s onwhitefl y mycosis enemies. For example, volatiles that are produced by spider mite-damaged plants of four gerbera cultivars differed in chemical composition and attractiveness to the predatory mite Phytoseiuluspersimilis (Krips, 2000).Th e host plant is also known to affect the efficacy of entomopathogenic fungi (e.g. Gallardo etal., 1990 ;Haje k etal., 1995) .Chemical s atth elea f surface can influence viability of fungal spores (Cooke & Rayner, 1984), but chemical composition may also have an effect on the pest species and its susceptibility to pathogens (Hare &Andreadis , 1983;Ramosk a &Todd , 1985).I n addition, with anincreasin g amount of trichomes on agerber a leaf, the searching activity of P. persimilis was hampered, aswa s itspredatio n rate of spider mites (Krips etal., 1999).Als oEncarsiaformosa, aparasitoi d of whitefly species, is influenced by trichome density on different cucumber cultivars (van Lenteren etal., 1995).Morphologica l factors also define the microclimate atlea f surface, in whichfunga l processes takeplace ,throug h leaf size, density of trichomes on theleave s and croparchitectur e (Ferro &Southwick , 1984).T owha t extent the architecture of ahos t plant is able to influence the efficacy of entomopathogens is still unknown. Here, we report on glasshouseexperiment s withA. aleyrodis andV. lecaniifo r thecontro l ofT. vaporariorum in gerbera carried out intw omorphologicall y andpossibl y also chemically different cultivars. Materials &Method s Whitefly-stockrearing Trialeurodesvaporariorum wa s reared on gerbera plants (Gerberajamesonii hybrids) for many generations over several years, in screened cages at 21°C (±2 °C) under greenhouse conditions. Gerbera cropand glasshouse conditions In a glasshouse of 150m 2tw o gerbera cultivars were grown: 'Bourgogne' and 'Bianca'. The Bourgogne cultivar produced red flowers and had 5-7 leaves per plant at the start of the experiment; 'Bianca' produced white flowers and had 8-15 leaves per plant. The leaves of 'Bourgogne'wer erelativel y smallcompare d toleave s of 'Bianca':ca .3 5x 10c mversu sca .4 5 x 16 cm respectively, but leaves of 'Bianca' were rather smooth compared with leaves of 'Bourgogne': 105trichomes/cm 2 versus ca.73 0trichomes/cm 2 onth eundersid e of the leaves (Kripset al., 1999) .Th enumbe r of abaxial stomata/mm2varie d for 'Bourgogne'betwee n 70- 75an dfo r 'Bianca'betwee n 65- 68 . The plants were cultivated on rockwool and were provided with water and nutrients viadri pirrigation , comparable to commercial growing systems.Th e cultivars were grown in 9rows ,whic h weretw oplant s wide and 29plant s long.Ever yro w was divided in five plots of 10( 2x 5 )plants ,whic hwer e separated bya nope n space( 1plan twide) .O never y sidefou r 115 Chapter 8 (2x 2)plant swer e left asbuffer , as well ason ero w atth e beginning, near the entrance.Th e eightrow swer edivide di n4 block si nnorth-sout h direction.Ever ybloc kconsiste d ofon ero w 'Bianca' and onero w 'Bourgogne'. The temperature and relative humidity (RH) in the glasshouse were measured every 5 minutes andaverage d overon e hour periods.Glasshous e temperature waso n average 19.2 °C (± 2.5°C SD), however on bright sunny days temperatures could rise up to 29 °C. The average humidity was 61.1% RH ± 14.1% SD,bu t during the second half of the experiment the RH fluctuated considerably (Fig. 8.1), between 20.2% and 80.4% RH (minimum and maximumdurin g 1 hour),a sa resul to fth elo whumidit ylevel soutsid ewhic hwer eo naverag e 30%RH .Onl ydurin gth eda yth emistin g systemwa soperated , sotha t thecro p would notb e wetdurin gth enigh tan dgrowt h ofplan t pathogenic fungi was avoided. This led to adiurna l rhythm of higher RH during the daytha n during the night. average day night 90 80 2 70 | 60 I 50 I 40 30 28 26 O 24 § 22 8 20 E ,2 18 16 75 80 85 90 95 100 105 110 115 Julianday s Figure 8.1: Average day and night temperatures (°C) and humidity levels (%RH ) in glasshouse; ->: indicates time of treatment. 116 Cultivar effects onwhitefl y mycosis Whitefly infestation in glasshouse Oncepe rtw oweeks ,7, 5an d3 week sbefor e thefirs t treatment, ca.300 0whitefl y adults,bot h male and female, were distributed over the gerbera plants. Every week the number of adult whiteflies perplan twer ecounted , starting4 week sbefor e thefirs t treatment.However , during thesepre-assessment s ofth ewhitefl y population, the cultivar Bianca seemed more attractive than the cultivar Bourgogne: young leaves of 'Bianca' contained up to 40 adults per leaf, whereas 'Bourgogne'ha d almost none.T o obtain aneve n distribution among thecultivars ,5 to 10adult so fbot hsexe swer econfine d toclip-cage s (0 2cm) ,on eclip-cag epe r'Bourgogne ' plant.Th eclip-cage swer eremove d after 24hr san d carewa stake nno tt odistur bth eadults . Thisprocedur e wasrepeate d once aweek , for 4weeks . Fungi Aschersonia aleyrodis originated from Aleyrodidae in Colombia (W. Gams, CBS, The Netherlands) and was passaged through whitefly twice. Sporulating colonies from pure cultureso npotat odextros eaga r (PDA,Difco ) wereuse d toinoculat emillet-culture s for mass productiono fconidi a(Fransen , 1987).Culture swer eincubate da t2 5° Can dL16:D 8 (artificial light). Conidia from 2-3 week old cultures were harvested by rinsing cultures with sterile demineralisedwate ran dsuspension s weredilute dt o 1 x 106o r 1 x 107conidia/m lwit h0.05 % (v/v)Twee n 80solutio n(Merck) .Initia lviabilit yo fconidi ao nwate raga r (15g/ 1 agar,Merck ) after 24hr s at25° C exceeded 95% germination. V. lecanii KV01 was obtained as commercial formulation Mycotal® (Koppert Biological SystemsBV ,th eNetherlands) ,whic hcontain s 1010 conidia/gram.Th eproduc t was suspendedi n sterilewate rt oobtai n 107conidia/ml , of which0. 1m l wasplate d onPD A and incubated at 23°C for 10 days. Subsequently, a conidial suspension was made, filtered (Wattmann filter no. 1) and diluted to 1x 106o r 1x 107conidia/m l with 0.05% Tween 80 solution.Initia lviabilit yo fV. lecanii conidia exceeded 98% germination after incubation for 24hrs .a t2 5 °Co nwate r agar. Treatmentprocedure Fivedifferen t treatments wereused , allotted toth e different plotswithi n aro w (4 replicates percultivar) .Th efung i ofbot hspecie swer eapplie d attw oconcentrations : 1 x 106 and 1 x10 7 conidia/ml and one plot per row was sprayed with 0.05% Tween suspension (control treatment). The suspensions were applied with a high volume hand sprayer (nozzle 55, pressure 4 x 105 Pa). Care was taken to hit the underside of the leaves and 'run-off of suspensionswa savoided .Th eamoun to f spraysuspensio n was 13001/ha(converte d from 150 m2) and 1.3 x 1012 to 1.3 x 1013 conidia per ha. This is comparable to amounts used in 117 Chapter 8 commercial practice(Wraigh t &Carruthers , 1999).Th etreatment s were applied once awee k in three consecutive weeks. Four weeks after the last treatment 40 leaves per plot were harvested and examined for total number of dead nymphs and overall number of whitefly nymphsan dpupa lcase spresent .Fo reac htreatment ,th enumbe ro fdea dnymph swer edivide d intw ocategorie s(i )mycose dnymph s (orangenymph s [A. aleyrodis] oropaqu ewhit enymph s [V. lecanii]) and (ii) desiccated nymphs (no apparent sign of infection: "mortality cause unknown") Whitefly developmenttime The development time of T. vaporariorum was monitored for the cultivars 'Bianca', 'Bourgogne'an d"Fame' .'Fame 'wa suse dt ob e ablet o compareth eresult s ofthi s experiment toresult so fothe rstudie s(Dorsma n& Vrie , 1987;Roermun d& va nLenteren , 1992;Siitterlin , 2000) and had ca. Ill trichomes/cm2 (Siitterlin &va n Lenteren, 1997). Six plants of each cultivar were grown in pots and placed together on ebb and flow tables in a glasshouse compartment (30m 2a t2 0± 2 ° C(SD )an d6 5± 10%RH) .Plant s werethre e months oldan d had atleas tfiv e leaves,o fwhic h one leaf wasused . For the experiment 50whitefl y adultso f both sexes wereconfine d in clip cages (0 2cm )o nyoung ,completel y unfolded leaves.Th e females weregive nth eopportunit yt ooviposi tfo r 24hr s andwer eremove d afterwards. Every one or two days egg hatching (day 2 to 17), adult emergence (day 28 to 42) and nymphal mortality were scored, until 80t o 90%o f the adults had emerged. Statisticalanalysis Fractions whitefly mortality in the glasshouse experiment were arc-sin square-root transformed. Transformed totalmortalit ywa sanalyse dusin ga two-wa yANOV Awit h (fungal) treatment and host-plant cultivar as independent variables, and blocks as block treatments (Genstat 5 release 3.22, Payne et al., 1987). Transformed "mortality by infection" and transformed "mortalitycaus eunknown "wer eanalyze di nth esam eway .Test swer e performed ata probabilit y levelo f 0.025,sinc eth edat ao ftota lmortalit yo non ehand , and "mortalityb y infection" and "mortality cause unknown" on the other hand, are linked. The subsequent differences werecompare d witht-tests ,i nwhic hprobabilit y was adjusted to thenumbe r of t- tests (n)performe d (0.025/n). The median hatching time and median emergence time were calculated from the hatching of nymphs and emerging of adults over time.A s every leaf is evaluated overtime , these hatching and emergence data were not independent. A separate regression analysis againsttim ewa stherefor e carriedou tfo r eachlea fi nrelatio nt ocultiva r(genera llinea rmodel , binomialdistribution ,logi tlin kfunction ) andth e50 %hatchin gan d50 %emergenc etim ewer e 118 Cultivareffect s onwhitefl y mycosis calculated. These values were compared using a one-way ANOVA with cultivar as independent variable (Genstat 5releas e 3.22, Payne etal., 1987) . Results Glasshouse experiment Inbot h cultivarsth efunga l treatmentscause d significantly highertota l mortality compared to thecontro ltreatmen t(Fig .8.2) .I ngeneral ,th ehighe rth econcentratio n ofconidi aapplied , the higherth ewhitefl y mortality. On 'Bianca', both fungi caused awhitefl y mortality upt o80 % (Fig.8.2) . On 'Bourgogne'V. lecaniicause d asignificantl y lower total mortality compared to A. aleyrodisan d compared to 'Bianca'. For both fungal treatments more nymphs on 'Bianca' died asfourt h instar/pupa than on 'Bourgogne' (Fig. 8.2). 'Bourgogne' was less attractive to whitefly adults than 'Bianca', therefore different whitefly infestation strategies were applied (see materials & methods). This could have resulted in a younger whitefly population on 'Bourgogne'. 100 [ 'Bianca' 'Bourgogne' 80' be 60 "5 40 / • . / •S 20; Tween 10 10 10 10 Tween 10 10 10 10 V.lecanii A. aleyrodis V.lecanii A. aleyrodis mortalityN1-N 3 ^n mm mortality N4-Pupae PZ3 ^M Control V.lecanii A. aleyrodis Figure 8.2: Average total mortality in gerbera cultivars Bianca and Bourgogne in control treatment (Tween 0.05%, white) or caused by V.lecanii (light grey) andA . aleyrodis (dark grey) applied in two concentrations: 106 and 107conidia/ml . Total mortality is divided according to the development stage in which mortality took place:unifor m = first tothir d nymphal instar, shaded =fourt h instar -pupa . Bars followed by the same letter are not significantly different, lsd = 0.178 (lsd is only valid when data are arcsW(fraction) transformed). 119 Chapter8 When mortalityb yinfectio n andmortalit yb yunknow ncause s were tested separately, the cultivar differences were even more pronounced. On 'Bianca' the infection levels were comparable to the total mortality levels (Fig. 8.2 and 8.3-1),wherea s on 'Bourgogne' amor e pronounced difference between the two fungi in infection levels was found. BothV. lecanii treatmentswer esignificantl y lowertha neithe rA .aleyrodis treatmen t (Fig.8.3-1) . A significant interaction between treatment and cultivar was found, indicating acultiva r influence onth e outcome of the treatment data. Mortality byunknow n causes showed a significant difference between thecultivars . On 'Bourgogne' the whitefly mortality by unknown causes was significantly higher than on 'Bianca' (Fig. 8.3-II) and the majority of these nymphs died in their first to third instar (data not shown). This effect is not very clear in the control treatments, as the differences are masked by accidental infection. This infection was caused by drift of conidial suspensions during spraying and was correlated with the treatments applied in the connected plots. Tween [•' • | V. lecanii A. aleyrodis 6 7 Tween 10 io 10 10 Tween 10 10' 10 10 V.lecani i A.aleyrodi s V.lecani i A.aleyrodi s Figure83 : Average whitefly mortality ingerber a cultivars Bianca and Bourgogne in control treatment (Tween 0.05%)o rcause db yA .aleyrodis an dV. lecaniiapplie d intw o concentrations. Mortalityi sdivide d inI :Mortalit y byinfection ; II:Mortalit y by causes unknown. Bars followed by the same letter are not significantly different; typographically different symbols indicatedifferen t setso f comparisons.Comparison s can onlyb emad e between different treatments within the same cultivar and between the cultivars for the same treatment; I: lsdaci = 0.256; II: cultivar effect: lsdAB = 0.234, treatment effect (*) lsd = 0.167: Tween -ab, V.lecanii 106- ab and 107- b, A. aleyrodis 106- ab and 107- a; lsd is only valid when data are arcsin\/(fraction) transformed. 120 Cultivar effects on whitefly mycosis Mortality by unknown causes was significantly higher forV. lecanii(cone . 107conidia/ml ) compared with A. aleyrodis(cone . 107conidia/ml) , but not significantly different from the control treatment (Fig. 8.3-II). Development time whitefly Thenumbe ro fegg sdeposite db yth efemale s inthi s 'nochoice ' situationwa sno t significantly different for the three cultivars (p= 0.465) . Also the development time of T. vaporariorum from egg deposit to 50% hatching (p = 0.425) and the median adult emergence time (p = 0.267)wer eno t significantly different for thethre e gerbera cultivars (Tab. 8.1). Thetota limmatur emortalit yfo r 'Bianca'wa so naverag e9.2% ,wherea sfo r 'Fame'an d 'Bourgogne' this was 11.6% and 13.8%respectively , but this was not significantly different (p = 0.564). Immature mortality was calculated using dead nymphs still present on the leaf surface. On 'Fame' and 'Bourgogne' many more nymphs went, however, missing than on 'Bianca'(Tab .8.1) .Thes enymph scoul dhav emove d outo fth eobservatio n area,whic h might haveresulte d inhighe rmortalit ythere .I twa sno tpossibl e to observe thewhol eleaf , because thebioassay swer ecarrie dou to nintac tplant s andleave seasil ybecam edetached .I naddition , dead nymphs will not be visible inevitably. On 'Bourgogne', for instance, more crawlers got stucki nth etrichome s and subsequently died than on 'Bianca' (Meekes,pers .obs.) ,bu t these wereno t traceable anymore after afe w weeks. Table8.1 : Number of eggspe r leaf and development time (days) from egg deposit* to 50% egg hatching or to 50% adult emergence of Trialeurodes vaporariorum on different gerbera cultivars at 20°C number of 50% hatched 50% adult total immature eggs per leaf eggs [days] emergence [days] mortality [%] missing Gerbera cultivar ± SD ± SD ± SD ± SD nymphs[% ] 'Bianca' 140 ±24 a 9.9 ± 0.2 a 35.9 ±1.3 a 9.2 ± 5.9 A 0.3 'Bourgogne' 135 ±33 a 9.7 ± 0.4 a 37 ±2.5 a 13.9 ±7.3 A 44.1 'Fame' 168 ±52 a 9.8 ± 1.5 a 35.6 ±3.2 a 11.6 ±2.5 A 48.3 : Egg laying completed within 24 hrs. Discussion Entomopathogenic fungi are able to cause up to 80% whitefly mortality in a gerbera crop grownunde rsemi-commercia l conditions.Th eresults ,however , showedtha tgerber acultivar s influence theefficac y of entomopathogenic fungi, depending on which fungal pathogen was used. Both fungi performed equally well on 'Bianca', whereas on 'Bourgogne' A. aleyrodis causeda highe rwhitefl y mortality thanV. lecanii(Fig . 8.2 and 8.3). These differences could 121 Chapter 8 bea resul to fdirec tinteraction sbetwee nfungu s andcultivar , orindirec t interactions withth e insect as intermediate. The host-plant cultivars differed inplan t architecture, leaf size and trichome density on aleaf , and probably alsoi n leaf chemical composition (Krips,2000) .Th e morphological features can influence the humidity at leaf level (Burrage, 1971; Ferro &Southwick , 1984; Chapter6) . 'Bourgogne'ha sfewe r and smaller leavestha n 'Bianca', whichcoul d enhanceth e exchange of water vapour on 'Bourgogne'. On the other hand, 'Bourgogne' has seven times more trichomes per leaf area than 'Bianca' (Krips etal., 1999), which could counteract the enhancedexchang e(Ferr o& Southwick , 1984).However ,no tonl yth eamount ,als oth eshap e of trichomes can differ substantially, e.g. single erect trichomes versus trichomes which entangleeac hother ,formin g awe bove rth elea f surface (Sutterlin &va nLenteren , 1997).Th e amounto ftrichomes ,thei rshap ean dth ecompositio n ofepicuticula r waxes alsoinfluenc e the hydrophobicity ofa lea f (Holloway, 1970).Leave s withfe w trichomes (like 'Bianca') enhance wetting,wherea sleave swit hman yweb-formin g trichomes(lik e'Bourgogne' )produc ea highl y waterrepellen t surface (Holloway, 1970).O nleave s of 'Bianca' droplets tend tob e larger and merge more easily than on 'Bourgogne', indicating that this might be the case. The higher wettingpossibilit y combined with adense rcro ppossibl yprovide d amor e suitableclimat e for entomopathogenic fungi on 'Bianca' than on 'Bourgogne'.Thi s would reduce the efficacy of V.lecanii o n 'Bourgogne'mor etha n A. aleyrodis,sinc eV. lecanii is more sensitive tolowe r humidity levels (chapter7) .O n 'Bourgogne',th ehighe r"mortalit y byunknow n causes"fo r V. lecanii (Fig 8.3-II)compare d toA. aleyrodiscoul d indicate that therelativ e humidity at the leaf levelwa sno t suitablefo r thefungu s toprotrud e outsideth einsect' sbody , norsho w signs of infection inside the host's body (white discoloration). When no protrusion takes place, infection byA. aleyrodisi s easier to observe, since infected nymphs turn an orange colour, whereasnymphs ,especiall y theyoun gones ,infecte d byV. lecaniima yno t alwaystur nwhite . Withincreasin ghydrophobicit yth edrople tretentio ndecrease s(Branskill , 1956).Sinc e the web-forming trichomes on a 'Bourgogne' leaf produce ahydrophob e surface (Holloway, 1970), the amount of inoculum present on a 'Bourgogne' leaf may be less, or is not as uniformly distributed compared with a 'Bianca' leaf. Besides, nymphs at the leaf plane are shieldedb yth elea f trichomeso n'Bourgogne 'an dles sconidi ama yreac hthem .Sinc e whitefly nymphs are sessile, but for a part of the first development stage (crawlers), direct hit of nymphs isth emos timportan trout eo finfection . The "lesser coverage"o n 'Bourgogne'shoul d affect bothfung i equally,bu tA. aleyrodisconidi aar efa rmor epersisten t (half-life of>2 5 days, Meekeset al., 2000)tha nconidi ao fV. lecanii (half-life of4 day so nchrysanthemum , Gardner etah, 1984).Thi senable sA. aleyrodist oprofi t morefro m asecon drout eo f infection viz.vi a sprayresidu e(Meeke set al, 2000).I ngenera l iti s suggested that, after hatching ofth eeggs , 122 Cultivar effects onwhitefl y mycosis crawlersmov eonl ya fe w millimetres before settling on asuitabl espo t onth eleaf ,wher ethe y remainfo r theres to fthei rnympha l development (Byrne& Bellows , 1991).O nbot h gerbera cultivarsnymph smove dconsiderabl y more- u pt oman ycentimetre sfro m theovipositio n site -,indicatin g thatthe yha d ample opportunity to come intocontac t with sprayresidue . Differences between cultivars may also affect entomopathogenic fungi via the pest insect e.g. development time or insect susceptibility. Development time and fecundity of T.vaporariorum differs between gerbera cultivars (Sutterlin, 2000) andth e faster theinsec t develops, the smaller the time frame will be in which the insect can be infected (Clancy & Price, 1987).N odifference , however, indevelopmenta l timean dnumbe r ofegg sbetwee n the cultivarsuse di nou rexperiment swa sobserve d (Tab.8.1) .Nevertheless , theseparameter s do not include preference of whitefly adults for one cultivar over the other. In the glasshouse experiment, whereadult sdi dhav ea choice ,'Bianca 'wa spreferre d over'Bourgogne' .Also ,th e "disappearance" of nymphs from cv's Bourgogne and Fame versus Bianca suggests that gerbera cultivar did affect the whitefly population. The preference of whitefly for 'Bianca' required adifferen t infestation strategy (seematerial s &methods) , which led eventually toa younger whitefly population on'Bourgogne '(se eFig . 8.2).If , for instance, alarge rpar t of the whitefly population on 'Bourgogne'stil lconsiste d of eggs, which areno t susceptible to either fungus, the chance that ahatche d nymph will get infected byA .aleyrodis i shighe r thanb y V.lecanii, duet oth elonge rpersistenc e ofA .aleyrodis conidia. Insect susceptibility to fungal pathogens is also known to be affected by host-plant differences in secondary metabolites (Costa &Gaugler , 1989;Gallard o etal., 1990;Poprawsk i etal, 2000).However , sofa r no secondary metabolites that may affect fungi through the pest insect have been reported on gerberaleave s (Harborne etal., 1999) .Althoug h various gerbera cultivars differ with respect tothei rchemica l composition, for instance,spide rmite-induce d volatiles of various cultivars differed inchemica l composition andattractivenes s toth epredator ymit eP. persimilis (Krips, 2000),i ti sunclea r atth emomen t if thisca n influence entomopathogens. Theabov eresult ssho wtha twhitefl y populationsi ngerber aca nb esuppresse db yusin g entomopathogenic fungi in aglasshous e situation undercondition swhic h arecommo n during commercial gerbera production (19 °C and 65%RH , Kerssies, 1994). Cultivar differences seem to be important: both fungi performed equally well on 'Bianca', but in 'Bourgogne' A. aleyrodiscause d ahighe r whitefly mortality thanV. lecanii. These differences seem tob ea resulto ffunga l characteristicsi nrelatio nt ophysica lplan tfactor s and/orinsec tcharacteristics , indicating thatA. aleyrodisi sbette r equipped todifferentia l circumstances thanV. lecanii. 123 Chapter 8 Acknowledgements Wear egratefu l for thetechnica l support ofJ .Tolsma .Specia l thanks weow e toJ . Meyvogel andH .Koedij k forth eexcellen t careo fth egerber a crop andJ.Th.W.H .Lamer sfo r adviceo n statistical analysis. We thank R. van der Pas (Koppert Biological Systems, NL) for kindly providing Aschersoniaculture s for our research. We also thank the E.A.M. Beerling (PBG, NL),th eEntomolog yPh.D .grou p(WU ,NL ) andH.R .Zoome rfo r criticallyreviewin gearlie r versions of this paper. This research was supported by the Dutch Product Board for Horticulture (Productschap Tuinbouw). 124 9 General discussion Microbialcontro lo fwhitefl y withAschersonia spp .comprise sman yaspects ,o fwhic hsevera l are addressed in this thesis.Th e previous chapters willb e summarized and discussed in the framework of the topics for evaluation of entomopafhogenic fungi for the control of insect pests (chapter 1,Tab . 1.3). Furthermore, theimplication s forAschersonia spp. as microbial control agentsfo r thesilverlea f andgreenhous e whitefly are given and suggestions for future research aremade . Fungal characteristics Hostspecificity and virulence Thespecificit y ofentomopathogeni c fungi hasbee n described as"th eexpressio n of reciprocal adaptations andaffinitie s between apathogeni corganis m andth eentiret y of itshos t species" (Fargues &Remaudiere , 1977).However , literature on entomopathogenic fungi concerning specificity doesno tprovid ea basi s todiscer n agenera lprinciple ,viz. som efunga l isolates are specific toon einsec tspecie so rgenus ,other s show highvirulenc e against specieso f various insect orders (Ferron etah, 1972;Fargues , 1976;Mo r etal., 1996;Vida l etal., 1997).I t is difficult tomak e astatemen t on specificity and virulence for the genusAschersonia o nbasi s of the results in chapter 2. The question whether someAschersonia species are specific to whitefly only and others to soft scales only, as wasdescribe d byPetc h (1921),ha s stillt ob e answered. Allisolate s originating from Aleyrodidae were ablet o infect Bemisia argentifolii and Trialeurodes vaporariorum, so it seems that those isolates have similar virulence characteristics.However , of 22isolates ,ou t of atota l of 44 isolates studied, theorigina l host insectwa sunknow no rwa sno tidentifie d tofamil y level,i nthi scas eCoccida eo rAleyrodida e (Tab.2.1 , chapter 2).Nin eou to fthes e2 2isolate swer eunabl et oinfec t eitherwhitefly . These isolatescoul db e specific for soft scales,especiall y since oneo f these isolates is A. turbinata, which has been described from coccids. It is possible that some species and/or isolates of Aschersonia may have broader host ranges than others and than previously was presumed (Petch, 1921), for instance, A. aleyrodis and A. placenta have been described from many Aleyrodidae species,bu tals ofro m Coccidae and occasionally Diaspididae (Tab. 1.2,chapte r 1). The morphological characteristic on which Aschersonia spp. are described as whitefly pathogenso rsoft-scal e pathogens, seemst ob ephenotypicall y determined, andtherefor e not areliabl etaxonomi c feature (Evans &Hywel-Jones , 1997).Th e aspect of taxonomy requires anurgen t update, inwhic h onema yconside r testing against arang e of possible host insect. 125 Chapter 9 Selectiono fa noptima l straini scrucia lt osuccessfu l application of entomopathogenic fungi (Hajek &St.-Leger , 1994).T oevaluat eentomopathogeni c fungi for thecontro l ofinsec t pests, severalfunga l characteristics canb econsidere d (Tab. 1.3, chapter 1).Hig h virulenceo f thepathoge n towardsth etarge t insect species is one of the most important selection criteria. Afunga l entomopathogen used asa biocontro l agent of whitefly in glasshouses should show highvirulenc eagains tB. argentifolii aswel l asT. vaporariorum, sincebot hwhitefl y species can occur at the same time. In general, the virulence of the different Aschersonia isolates againstB. argentifoliiwa spositivel yrelate dt oth evirulenc e againstT. vaporariorum (chapter 2, Fig. 9.1). Several Aschersonia isolates were not only highly virulent towards both whiteflies, but also showed a high spore production on semi-artificial medium, another prerequisite for microbial control. Figure 9.1: Relationbetwee n total mortality and mortality by infection of silverleaf + mortalityb yinfectio n whitefly, B. argentifolii, and greenhouse totalmortalit y whitefly, T. vaporariorum, by eleven isolates of Aschersonia spp. (see also Fig. 2.5, chapter 2). Correlation coefficients: 20 60 80 100 mortality by infection r2 = 0.92, total % mortality of silverleaf whitefly mortality 1^ =0.91 . Medianletha ldose sfo rfiv eisolate so fAschersonia spp.varie dbetwee n 1650t o477 0 conidia/cm2forZJ .argentifolii, andbetwee n 710t o 5340conidia/cm 2fo r T. vaporariorum on poinsettia (chapter 3).Thes e values are higher than those found by Vidal et al. (1997) for Paecilomycesfumosoroseus, but lower than the values Wraight et al. (1998) found for B. bassiana,P. fumosoroseus and P.farinosus (see also chapter 3).Regressio n coefficients (probit-slopes) of these five Aschersonia isolates varied between 0.82 and 1.32 for B. argentifolii andbetwee n 0.83 and 1.44 forT. vaporariorum, which isi n the samerang ea s thecoefficient s found byVida l etal. (1997 ) andWraigh t etal. (1998) .Fo r fungal dose-host mortality responses these coefficients typically vary between 0.5 and 1.5 (Wraight & Carruthers, 1999),an d arelo w compared to chemical compounds.I n this thesisA. aleyrodis 126 General discussion andA .placenta showe da highe rvirulenc etoward sT. vaporarioruman dgav emor econsisten t resultstha nV. lecanii (chapter6, 7 and8) ,althoug hMead e &Byrn e (1991)an d Gindin etal. (2000) found high infection levels of B. argentifolii and T. vaporariorum by V. lecanii. Unfortunately theseresult s aredifficul t tocompar e sincen oLD 50value s aregiven . Athig h doses,conidi ao fAschersonia spp. seem to posses self-inhibitors (chapter 3). These self-inhibitors could be advantageous for the fungi, to prevent germination during conidiogenesis andgerminatio n enmasse. Iti sunlikel y that this will interfere with microbial pestcontrol ,sinc ethi sphenomeno n onlyoccur sa tver yhig h dosages (1.4x 105conidia/cm 2), whichar eunnecessar y anduneconomica l for microbial control. Median lethal time of the five Aschersonia isolates varied between 5 to 7 days for B. argentifolii, and5 an d 10day sfo rT. vaporariorum (chapter 3).Sinc ewhitefl y nymphs are sessile,th eexac ttim eo f death isrelativel y difficult todetermin e compared tomobil einsects . TheLT 50value s aredetermine d usingsign so f infection, sucha sdiscolouration , will probably underestimateth eactua ltim eo fdeath ,becaus eactua ldeat hwil lhav etake nplac ebefor e signs of infection occur. When the nymphal stage is used to determine time of death, death will occur on average one nymphal stage later than the nymphal stage treated at 20 °C, which means that 5t o 7 days after inoculation 80-90%o f the whitefly population dies of infection byA .aleyrodis andA .placenta (chapte r 4). Thisi s faster compared to theLT 50value s found for theseAschersonia isolates (chapter 3). Bioassaymethod sdepen do nth econsistenc y ofth emetho d when comparing different isolates of fungal pathogens. In the case of B. argentifolii andT. vaporariorumth e use of intactplant sappear st ob ea suitabl esystem ,sinc econsisten tresult swer efoun d overtime .Th e bioassaydi dnee dsom eadjustin g compared toFranse n (1987),viz. Fransen(1987 )use d2 4hr s of highhumidit y for cucumber plants,wherea sth estandar dbioassa y usedi nthi sthesi s was 48 hrs of high humidity for poinsettia plants. It was necessary to create a more favourable climate, since poinsettia seems to have a more hostile environment for entomopathogenic fungi (chapter5, 6 and 7).Onc ea mor efavourabl e situationi s created on poinsettia LD50an d LT50value s of A.aleyrodis o n this crop (chapter 3) are comparable to results obtained with A.aleyrodis oncucumbe r (Fransen, 1987).However , sinceo npoinsetti amortalit y levels are rarely above 85- 90 %(chapte r 2- 4) ,th eLD 95value s willprobabl y differ. Although bioassays carried out on intact plants are more labourious than a bioassay performed on leaf discs and are not always feasible considering space and costs, many advantagesexist .Sid eeffect s likeprematur eagein go fleave sd ono t occur,leave ssta yi nthei r naturalpositio nan dsystemi ceffect s onal ltrophi clevel sca nb estudied . Sprayingleave swit h afunga l suspension,instea d ofdippin gleaves ,an dperformin g thebioassay sunde r glasshouse conditions makesth eresult s easier to extrapolate to large glasshouse applications. 127 Chapter 9 Sporulation, dispersaland persistence Whitefly nymphs infected with Aschersonia spp. may show extensive sporulation if the environmental conditions are favourable. However, onpoinsetti a sporulation only occurred inhumi d spotswithi n acrop ,wherea so ngerber ai twa softe n observed (E.T.M.Meekes ,pers . obs.). Even when extensive sporulation isobserved , secondaryinfectio n will belimite d and isno texpecte d tocontribut e toeffectiv e whitefly control in aglasshous e situation (Fransen, 1987).Conidi awhic har eproduce di nslim yhead susuall yar edisperse db ywate ro rb yinsect s (Samson &McCoy , 1983;Osborn e& Landa , 1992),althoug hAschersonia conidiamigh tals o besubjec t towin ddispersa li nnatur ewhe nfree d of theirmucilag e matrix anddrie d bywin d (Morrill &Back , 1912).Sinc ei tdoe sno trai ni n aglasshous e and overhead irrigation israre , only condensation of water onth e leaf surface may enable conidia tob e transported from a sporulating insect to a susceptible one on the same leaf. Transmission of conidia by other insects likeEncarsia formosa (Fransen &va n Lenteren, 1993) or mites (Osborne &Landa , 1992)wil lprobabl yno t contribute toa hig hrat eo finfectio n either,especiall y within theshor t timeperio davailabl efo r controlo f awhitefl y population ina glasshous ecro p(Fransen , 1987). In a glasshouse situation conidia of A. aleyrodis were able to stay viable for over a month andinfec t hatchingT. vaporariorum nymphs (chapter 5);60-80 %o fA .aleyrodis an d A. placentaconidi awer estil labl et ogerminat eo nwate raga rafte r 40day ssta yo na poinsetti a 100 Figure 9.2: Mean percentage germination ofA .aleyrodis an dA. placenta conidia from 0 days until 40 days on leaf surface of gerbera and poinsettia at 20 °C; by direct observation of agar imprint of leaf (= germination on leaf surface): - -V- -: A. aleyrodis,—•— : A. placenta, orobservatio n after 24hour sincubatio n onW A at25° C (= germination capacity): - - O - -: A. aleyrodis, —•—: A. placenta. Vertical lines represent standard error of mean (for more details see chapter 5). Tapaninen & persistence (days after inoculation) Meekes, unpublished results. 128 General discussion leaf surface (Fig. 9.2; Tapaninen & Meekes, unpubl.). Although Aschersonia spp. were evaluated for the control of whitefly species in glasshouses, where the environment is less extremetha n inope nfiel d situations, conidia ofAschersonia spp .ar e also likely topersis t in field situations. In (sub)tropical forest ecosystems, its natural environment, where bright sunshine alternates with clouds and rain, Aschersonia spp. have to overcome extreme circumstances.Conidi ao fman yAschersonia spp.contai npigments ,fo r instanceA. aleyrodis conidia contain carotenoids (Eijk et ah, 1979), which may enable them to withstand short periods of exposure to radiation better than hyaline spores, although variation between and within species exists (Fargues etal., 1996).Moreover , conidia are directed against whitefly nymphs which are mainly present atth eunderside s of leaves and thus protected from direct sunlight. Alsoth emucilag ematri xo fconidi ama yb eabl et oac ta sa nantidesiccan t (Boucias & Pendland, 1991) and protect the conidia against environmental stress (Louis & Cooke, 1985). Thelon gpersistenc eo fAschersonia conidi ai sprobabl ya phenomeno n thati scommo n withinth egenu sAschersonia, sincespecie so fAschersonia an d their teleomorphHypocrella werecapabl et owithstandin glon gperiod s of drought inthei rnatura lenvironmen t (upt o 5-6 months)(Evan s& Hywel-Jones ,1997) .Th eabsenc eo fgerminatio n atth elea f lamina(chapte r 4 and5) ,ma yals ob ea strateg yt osta yviabl eove ra prolonge d period.However , viabilityan d especiallyinfectivit y arehost-plan trelate d (chapter 5, 6 and7 )an dth eexten to f thisresidua l effect on whitefly control is dependent of whitefly distribution over the plants and plant growth. Cultureand mass production Ingeneral ,th emajorit y ofAschersonia spp.grow s and sporulates well on most conventional mycologicalmedi a and semi-artificial media, likebrow n rice,soybean s or millet (Ramakers etal, 1982;Samso n etal., 1982 ;Ramakers , 1983;E.T.M .Meekes ,unpubl .res.) ,bu t notable differences in nutritional requirements are apparent; some isolates had a higher conidial production on millet, whereas others sporulated better on corn medium (chapter 1, E.T.M. Meekesunpubl .res.) .Optimu m temperature for growth and sporulation ofAschersonia spp. varies between 25 to 30 °C (Spassova, 1980;Rombac h & Samson, 1982) and no mycelial growth isobserve d attemperature s exceeding 31- 3 5°C ,dependin g on the isolate (Rombach & Samson, 1982;Samso n &Rombach , 1985).Sporulatio n canb estrongl yenhance d by light, ofwhic hwavelength s betweenblu ean dblac kligh tar emos tfavourabl e (Samson etal., 1982 ; Rombach & Gillespie, 1988). After subculturing A. aleyrodis 12 times on millet, final infection levelso fT. vaporariorum were similart otha to f isolatestha twer e subcultured three times (Fransen, 1987), however, one has to be careful extrapolating these results to other Aschersoniaisolates .Mal textrac taga ri smos tfavourabl e forisolatio n andmaintenanc ei nth e 129 Chapter 9 laboratory, althoughgrowt hi s slower and sporulation is less abundant compared with media containing peptone (Rombach &Gillespie , 1988;E.T.M . Meekes, unpubl.res.) . On several other media (e.g. yeast/peptone/glucose), a sudden deterioration of colonies may occur and valuableculture sca neasil yb elos t (Rombach &Gillespie , 1988).Thi sdeterioratio n seems to be common among entomopathogenic fungi, therefore, regular passage (e.g. once a year) through a suitable host insect followed by subsequent reisolation, or suitable storage of the original isolate is recommended in order to maintain viability and infectivity (Kogan & Seryapin, 1978;Samso n &Rombach , 1985;Hirt e etal, 1989a;But t &Goettel , 2000). Themajorit y ofcontro lprogram sutilizin g fungi relyo ninundativ e releases (McCoy, 1990;Haje k &St.-Leger , 1994)an drequir elarge-scal e application ofth efungu s eachtim eth e pest population exceeds its action threshold. Small scale mass-production of the fungus alreadystarte d atth ebeginnin g of the 1900's,whe nA. aleyrodisan dA. goldiana were grown in wide-mouth one-pint bottles on sweet potato agar (Berger, 1920b; 1920a; 1921). In the 1970'si nEaster nEurope , severalAschersonia spp .wer emass-produce d onbee r wort with a sugarconten t of 10%,a t24-2 6 °Can d 80-100%R Hi nluminescen t light (Kogan &Seryapin , 1978;Spassova , 1980).Afte r 25-30days ,th eculture s couldb euse do rrefrigerate d (<4 °C ) for upt oon eyea r (Kogan &Seryapin , 1978).Th eadvise d fungal dosepe rh a varied between 1012 and 1014conidia , which could kill up to 98%o f the whitefly nymphs after three to four treatments (Kogan &Seryapin , 1978). Standard technology for mass production of several entomopathogenic fungi has employed asubstrat eo f cooked riceo r other grains ontrays ,i n autoclavablebag s ori n glass jars (Wraight & Carruthers, 1999). Cereals form attractive solid substrates, being cheap, nutritious, and easy to sterilize (Hall & Papierok, 1982). However, average yields of approximately 1-5 x 109conidia/ g dry weight of substrate are achievable, and is generally adequatet osuppor t application rates of 1 -5 x 1012conidia/h a(Wraigh t &Carruthers , 1999). Commercial scaleproductio n capacity necessary to support multiple applications tocrop sa t ahig hrat eo f 1013conidia/h aa tcost scompetitiv ewit hsyntheti cchemica l insecticidesha slon g represented amajo r production barrier(Wraigh t& Carruthers , 1999).Th estandar d technology used by industry for production of fungi and bacteria is submerge culture, by which the microbei sproduce d ina standardize d liquidmediu m (McCoyet al., 1988) .Harvestin g spores from submergedfermentation s ismuc heasie rbecaus eth e sporesca nb ereadil ycollecte d and concentrated by centrifugation or filtration and then dried (Bartlett & Jaronski, 1988). Aschersoniaspp .d ono t produce conidia in liquid culture, oronl y atit s surface (Hirte et al, 1989a; Ibrahim etal., 1993),whic h hasbee n a draw back for their massproduction . Many otherentomopathogeni c fungi, such asB. bassiana,P. fumosoroseus andV. lecaniiproduc e "blastospores"i nsubmerg ecultur e (Humphreys etal., 1986;Hirt e etal., 1989b ;Fen g et al., 130 General discussion 1994).Althoug hi n somecase svirulenc e of theseblastospore s ishighe rtha ntha to f conidia, theyar emor esensitiv et o adversecondition s (Hall, 1981;Shimiz u& Aizawa , 1988;Lace y& Mercadier, 1998) and therefore attention is shifted to production of conidia on solid phase systems. To massproduc eAschersonia spp .cost-effectively , two barriers have to be taken. Firstly, conidia areproduce d in mucilage which asks for adifferen t approach for harvesting conidia, in contrast to e.g. B. bassiana or P. fumosoroseus which produce hydrophobic conidia. Secondly, Aschersonia spp. need at least two to three weeks to reach an optimal production on semi-solid media, in contrast toB. bassiana,whic h starts sporulating within a few daysafte r inoculation of growth medium (Wraight &Carruthers , 1999).Eac h organism has its own particular requirements - substrate type, nutrients, etc. - and mass-production efforts mustb eindividuall y tailored for optimalyield s (Bartlett &Jaronski , 1988). Research intoth eoptimu m growth medium tomas sproduc eAschersonia spp.i sneeded , but therecen t interesti ntwo-phas e and solid-phase fermentation technology willoffe r new perspectives for future massproductio n (Fransen, 1990;Lace yetal, 1996;Wraigh t &Carruthers , 1999;But t & Goettel,2000) . Formulation Formulation cangreatl yimprov eth eefficac y ofentomopathogen s both inprotecte d and field crops. There are four basic functions of formulations: 1) to stabilize the organism during production, distribution and/or storage, 2) to aid handling and application, 3) to protect the biocontrol agent from harmful environmental factors and 4) to enhance activity of the organism (Jones &Burges , 1998).I ti s generally accepted that stability (shelf-life) for 12-18 months without refrigeration would be required to service general agricultural markets, although stability of 3-6 months would be sufficient for products produced on contract for applications at a specific time (McCoy, 1990; Wraight & Carruthers, 1999). The type of formulation andselectio n of additives dependso nth ebiolog y andphysica l properties of the pathogen, host insect and target crop (Soper &Ward , 1981).Th ebasi c components of most formulations include, in addition of fungal conidia, one or more of the following: a carrier, diluent,binder , dispersant, UVprotectant s and virulence-enhancing factors (Butt &Goettel , 2000). Additivestha thav ehygroscopi can dadhesiv epropertie sar eimportan tt oa formulation sinceth emod eo f action for fungi generally involves integumental contact of the hostb yth e infective unitunde rcondition so fhig hmoisture .A swit hothe rpesticides , wetting agents that reduceth e surface tensiono fth espra ydrople t areimportan t inmaximizin g distribution of the infective unit,especiall yconsiderin gth ehydrophobi ccuticl eo fth einsec thost s(chapte r4 )an d differences inepicuticula r waxlayer so f thevariou shos tplant s (Fig6.5 ,chapte r 6). Conidia 131 Chapter9 have been generally formulated as awettabl e powder ordus t for those fungi that havebee n successfully mass produced. In a wettable powder, conidia are generally diluted with inert fillers, wetting agents and spreaders. Water is often used as the carrier. Where conidia germinate quicklyi nwate r adus ti spreferred . Talc,flou r andmil k powderhav ebee n used as suitable diluentsfo r dusts (McCoy, 1990). Themos timportan trecen tdevelopmen ti nformulatio n of conidiaha sbee nth eus eo f oil carriers. Oils are relatively effective in sticking conidia toinsec t andplan t surfaces, thus eliminating the need for wetting, sticking, or spreading agents. The use of oils also circumvents thedus t problems associated with wettable powders and oils are more effective carrierso flow-volum eapplication s than water (Wraight &Bradley , 1996).I n addition,the y canenhanc efunga l performance atlo whumidit ylevel s (Batemanet al., 1993) .However , oils aremor eexpensiv ean dca nb ephytotoxic ,henc eshoul db eapplie d atlo wrates ,an dman yar e insecticidal themselves (Wraight & Bradley, 1996; Butt & Goettel, 2000). They do not necessarily enhance the performance of entomopathogenic fungi, since in field tests of B. bassianaagains tB. argentifolii emulsifiabl eoi lformulation s ofconidi aperforme d nobette r than aqueous suspensions prepared with a highly effective wetting agent (Wraight & Carruthers, 1999). Regarding Aschersonia spp. considerable research input is still needed before aproduc t willb e available to growers.A s with massproduction , everyfungu s hasit s ownparticula rrequirement s andth eformulatio n shouldno tinterfer e withth einfectio n process (Butt& Goettel ,2000) .However ,funga l formulation isprobabl yth eke yfacto r topas t failure andfutur e successes (McCoy, 1990). Safety Potential safety issueshav et ob e addressed when microbial biocontrol agents areintende d to control insect pests (or diseases). Unintended adverse effects mayb e fourfold (Cook et al., 1996): 1) Competitive displacement. Microorganisms could displace naturally occurring microorganisms throughcompetitio n for nutrients (Cooket al., 1996) .Sinc eAschersonia spp. originate from the (sub)tropical areas and are specialized on Aleyrodidae (Evans &Hywel - Jones, 1990),i tis ,therefore , unlikelytha tthe ywil lestablis h themselves in temperate regions outside the glasshouse and eradicate their host insect or displace other entomopathogenic fungi. 2) Allergenicity Certain kinds of pollen and airborne fungal spores are inevitably present in the air webreat h and cause allergies in sensitive people.Potentially , a biocontrol micro-organism released in the air could elicit allergic reactions (Cook etal., 1996). Since conidia of Aschersonia spp. are produced in mucus and conidia are normally not airborne (Evans & Hywel-Jones, 1990), this risk is relatively small. Thereby, only a very small 132 General discussion proportion of fungal species produce spores that cause allergies (Latge & Paris, 1991). In addition,an ymicrobia lproduc to fwhic hi sclaime dtha ti tsuppresse spest s(o rdiseases )need s tob eregistere d onth eDutc h marketunde rth epesticid e law.Suc hproduct s shouldb e applied with the necessary precautions (e.g. wearing protective clothing and mask). Exposure to allergenicparticle so f alltype si scommo n in agricultural settings,an dallergie s willno tb ea newproble mbecaus eo fth eus eo fbiocontro l agents,bu tshoul db eaddresse d asa safet y issue during development and application (Latge &Paris , 1991). 3) Toxicity of biologically active metabolites. Fungi secrete a wide range of metabolites, some of which are important medicines or research tools, some are toxic or carcinogenic (Strasser et al., 2000b). Chemically diverse, toxic metabolites have been described in several entomopathogenic fungal genera, such asBeauveria, Paecilomyces and Verticillium (Khachatourians, 1991;Gindi net al, 1994; Boucias, 1998;Strasse r etal, 1998), and also A. aleyrodisan d A. insperatasee m to produce exotoxins in vitroo n nutrient-rich media (Kurbatskaya etal., 1983;Krasnof f etal., 1996).Althoug h some of these secondary metabolites areknow nt ob eimportan t pathogenicity determinants, therol e of others remains unclear (Strasser etal, 2000a; 2000b).Ther e is concern that these toxic fungal metabolites may enter the food chain and form aris k to humans and animals. However toxin quantities produced invivo ar eusuall y far lower than thoseproduce d innutrient-ric h media, and inter- and intra-specific variation in the production of fungal metabolites exits (Strasser et al., 2000b). It is also important to put the relative amounts of toxins produces by entomopathogenic fungi intoperspective , since themajorit y of fungal toxins which enter the food chain in significant levels and frequency to cause concern are produced by common saprophytic fungi, like Aspergillus, Fusarium and Penicillium (Steyn, 1995). Entomopathogenicfungi ,unlik esaprophyti cfungi , dono tsecret emetabolite so nplan tmateria l in sufficient quantities to cause ahealt hris k (Strasser etal., 2000b) . 4) Pathogenicity. Microbial biocontrol agents used as pathogens could also be pathogenic tonontarge t organisms.However ,Aschersonia spp.ar eno texpecte d to cause any negativeeffect s toth eenvironmen tbecaus eo fthei rhig hspecificit y towardswhiteflie s and/or soft scales (Evans &Hywel-Jones , 1990) and areunabl e toinfec t beneficial insects. Compatibilitywith other control methods Theapplicatio n ofAschersonia spp .ha st ob eintegrate d into amanagemen t system of other controlmeasure sagains tpest s and diseases.Th eus eo f fungicides especially mayimpai r the efficacy ofthes efungi . Atth ebeginnin go fthi scentury ,Fawcet t (1908)alread yremarke dtha t applications ofA .aleyrodis conidiawit h sprayequipmen t whichwa searlie r usedfo r spraying pesticideso rwhic hcontaine dcopper ,wer eno tsuccessful . Occasionaltest sconcernin gmixin g 133 Chapter 9 of A. aleyrodisconidi a with fungicides and insecticides were made. Oho (1967) found that conidia did not germinate in solutions of any fungicide tested. However in reduced concentrations ofmaneb , germination was not affected. Zibul'skaya etal. (1975 ) found that sensitivity tofungicide s wascorrelate d withth eag eo f theAschersonia culture,culture s of 30 days old were less sensitive compared to cultures of 60 days old. Application of the detrimental fungicide mancozeb did not impair infection of third instar nymphs of T. vaporariorumwhe napplie dthre eday slate rtha nth espor esuspension , whenth efungu s had already infected the insect successfully (Fransen, 1987).Thus after a certain safety period fungicides can be applied. Some of the insecticides tested by Oho (1967) suppressed germination {e.g. dichlorvos), whereas the majority of insecticides had no adverse affect at their recommended dose. Also miticide Avermectin Bl was not harmful to A. aleyrodis (McCoy etal., 1982).Th e application ofAschersonia spp . together with new fungicides or insecticides still has to be tested. Considering the use of fungicides or insecticides together with entomopathogenic fungi, they should not actually be mixed together, but applications should be scheduled atcertai n time intervals. Thecombine d use ofnatura l enemies (parasitoids andAschersonia spp. ) of whitefly isfeasible .Fo rinstance ,E. formosa i sabl et odistinguis hinfecte d hostsfro m noninfected hosts (Fransen& va nLenteren , 1993).I nadditio nE. formosa prefers olderinsta r nymphs (Fransen &Montfort , 1987),whic har eles ssusceptibl et oA. aleyrodis andA .placenta (chapte r4) .Onl y whitefly hosts parasitized by E. formosa less than four days before inoculation with A.aleyrodis ma ybecom einfecte d andth eparasitoi dma ydi e(Franse n& va nLenteren , 1994). Duet othei rhig h specificity,Aschersonia spp.d ono tpos ea ris k toothe rnatura l enemies and beneficial organisms. Host characteristics Susceptibility First, second and third instar nymphs of both whitefly species are highly susceptible to A. aleyrodis andA .placenta (chapter 4), whereas whitefly eggs are not infected. Whitefly adults are rarely infected, but onlyi n an environment with 100%R H (Fransen etal., 1987) . Thefourt h instarnymph sbecam eles ssusceptibl eove rtim ewhe nthe ydevelo pint opupa .Thi s phenomenon isremarkabl y similar for B. argentifoliian dT. vaporariorum (chapter 4).Thi s maturation immunity of nymphal stages is acommo n feature, which hasbee n described for manyinsect-fungu s combinations (Tanada &Kaya , 1993). For many years it wasbelieve d that development of resistance topathogen s used for pestcontro l wouldb eunlikely . This wasbase d onpathogen s being living organisms that are continually coevolving with their host, unlike chemical pesticides, and have been able to 134 General discussion coexistwit hthei rhost s overeon s(Briese , 1986).However , resistance has developed in some combinationso fpes tan dbacteria l orvira lentomopathogen s and/or theirproducts ,bu t so far for fungal entomopathogens nocase s of trueresistanc e of target insects are known, although this mayb e due to the relatively limited use of entomopathogenicfung i ascontro l agentso r a lack of follow-up studies. There is considerable intraspecific variance in fungal species resulting in differences in pathogenicity, and likewise, there is simultaneously variability in the host (Shelton & Roush, 2000). Diversification of the flora in tropical forests led to increased specialization of plant-sucking homopterans and acoevolvin g host specialization of theirpathogen s (Evans, 1988).Althoug h pathogens, which arewel l adapted to their host, were considered to be relatively benign, this is not necessarily the case. A pathogen might evolve to relatively high virulence when its dependence on the infected host mobility is reduced by using other transmission modes, for instance, dispersal by water or wind. Pathogens may also reduce their reliance on host mobility by a 'sit-and-wait' transmission, whichmean srelianc eo nmobilit yo fsusceptibl ehosts .I ngeneral ,thes ecategor yo f pathogens are very persistent (Ewald, 1995).Thi s mayexplai n the high virulence ofAschersonia spp., which have to rely on other modes of transmission, since their host is sessile,bu t other host and pathogen characteristic are also likely to play a role.Wit h the application of microbial insecticides inglasshouse s onewoul dexpec t ahig h selection pressure,bu t sinceAschersonia spp.ar eno tendemi cpathogen s constantly present inth e whitefly population, this is unlikely to happen. Combined useof , for instance,th eparasitoi d E.formosa reduces this probability even more (Fransen, 1987).I n general, successful laboratory selection for resistance against fungal entomopathogens does not appear to have taken place, and there are no documented cases of resistance in thefiel d (Shelton &Roush ,2000) . Whitefly distribution andits implications Within aplant ,adult s ofT. vaporariorumprefe r young leaves for feeding and oviposition.A typical vertical distribution of development stages within the plant arises, because of the combination ofpreferre d oviposition sites,plan tgrowt han dnymph sbein glargel ysessil e(va n Lenteren &Noldus , 1990).Th edistributio n of thegreenhous e whitefly is largely aggregated, ona leaf ,withi n aplant ,betwee n plants or on glasshouse scale (depending onth e size of the glasshouse). Vertical distributions of B. argentifoliiwithi n aplan tresembl e those described for T. vaporariorum, but are probably less pronounced (van Lenteren & Noldus, 1990) probablydu et o their slower developmental time andresultin g overlapping nymphal cohorts (chapter4 ;Fransen , 1994; Drostet al., 1998) .Th edifferentia l susceptibilityo fth evariou s life stageso fwhitefl y (chapter4 )wil lrequir ereapplicatio n ofth efungu s toachiev ea goo dcontro l of a whitefly population over time. However, this also offers perspectives for the use of 135 Chapter 9 E.formosa, whichprefer s olderinsta rnymph sfo r oviposition (Fransen &Montfort , 1987),i n combination with thepathogen . Environmental influences Hostplant In this thesis it was shown that the host plant is an important factor in the efficacy of entomopathogenic fungi intritrophi csystems .I tinfluence s theefficac y of entomopathogenic fungi directly, e.g. by affecting persistence ofA .aleyrodis, an dindirectl yb yinfluencin g the developmentrat eo f insectand/o r theirsusceptibilit y aswel la sth ephyllospher ehumidit y and conidial deposit. Thepersistenc eo fA .aleyrodis conidi awa sdirecd yaffecte d byth ehost-plan tspecies . In the study on cucumber, the germination capacity and whitefly mortality exhibited by A. aleyrodis remained high over a period of at least one month, but on gerbera both germination capacity and whitefly mortality declined, suggesting a less favourable environmentfo rpersistenc eo fconidi ao ngerber a(chapte r5) .Thi scoul db edu et o differences in microflora, differences in chemical composition of the host plant, e.g. availability of nutrients of fungal-static compounds, and/or differences in phyllosphere humidity. On poinsettia nymphal mortality declined considerably in time, despite the high germination capacity of conidia (chapter 5). This could mean that conidia on poinsettia leaves are still viable, but lost their ability to infect, but it could also mean that low infection levels on poinsettiaar eno trelate dt ofunga l persistence.Beside sth edirec teffect s mentionedabove ,als o indirect effects via the host insect maypla y arole . Developmentalrate ,reproductio nan dmortalit yo fwhitefl y areinfluence d byhost-plan t species (van Lenteren & Noldus, 1990). The control mortality of greenhouse whitefly was significantly highero npoinsetti a relativet ocucumber , gerbera ortomat o (chapter 6; seeals o van Lenteren & Noldus, 1990). A less suitable host plant, as reflected by higher control mortality, couldresul t ina nincrease d susceptibility to entomopathogens, due to stress of the host insect (Steinhaus, 1958).I n contrast, this could also result in adecrease d susceptibility to entomopathogens compared to insects from a more suitable host plant, since these 'healthier' insectsbette rsuppor tgrowt h ofnatura lenemie s (Schultz &Keating , 1991;Hoove r et al., 1998). However, here low fungal infection is probably not directly linked to plant quality for the host insect, since B. argentifolii exhibited the same low infection levels on poinsettia asT. vaporariorum evenafte r 48hr shig hhumidit y (>95%RH ,Fig .9.3) , whileo n poinsettiaB. argentifoliishow sa lo wnatura lmortalit y(Fi g9.3 ;va nLentere n& Noldus ,1990 ; Fransen, 1994). Nevertheless, aberrant growth of germinated Aschersonia conidia was occasionallyobserve do nth ecuticl eo fB. argentifoliinymph s(chapte r4) ,whic hma yindicat e 136 General discussion Poinsettia Cucumber Tomato 1001— Tween A.aley . A.plac . Tween A.aley . A.plac . Tween A.aley . A.plac . 'other causes' infection mortality of B. argentifolii R^ [ j mortality of T. vaporariorum K\l WM Figure 9.3: Mean percentage mortality of third nymphal instars of B. argentifolii (dark grey bars) and T. vaporariorum (light grey bars) byA . aleyrodis andA .placenta (107 conidia/ml) and other causes (shaded bars)o npoinsetti a ('Goldfinger'), cucumber ("Profito') andtomat o CMoneydor')afte r 48hr shig hhumidit y (>95% RH, see also Materials and Methods of chapter 2 and 3);erro r bars represent standard error of mean. possible lower susceptibility in apar t of thehos tpopulation . Theepicuticle ,th efirst barrie r thefungu s hast obreach ,contain slarg e amountso ffre e lipids,whic h areabl et oinfluenc e the fungus, byfo r instanceinhibitin go r stimulating germination orformatio n of non-penetrating germtube s (Boucias& Pendland , 1991).Fo rB. argentifoliith ecompositio n of theselipid s is directly related to host-plant species, but for T. vaporariorum this composition is mainly determined byth einsec t itself andonl ypartl yb yth ehos tplan t (Nealet al., 1994) . The development rate of T. vaporariorumwa s highest on cucumber and lowest on gerbera,however ,difference s at2 0° Car e small (ca.4 daysbetwee n 50%emergenc e adults, E.T.M.Meekes ,pers .obs.) .Hence ,th echance s of escaping fungal infection, which involves alsoth elos so finoculu mb ymoulting ,ar eabou tth esam efo r whitefly nymphso nth e different hostplant s duet oequa lexposur etime s (window of opportunity;Clanc y &Price , 1987).Thi s is a general phenomenon in many insect -funga l pathogen interactions (Ferron, 1985).Th e cotton aphid strainuse di nchapte r 6,whic h didhav e ahighe r development and reproduction rateo ncucumbe rtha no ngerber a(E.A.M .Beerling ,pers .obs.) ,i smor elikel yt oescap e fungal infection on cucumber (Clancy & Price, 1987). In addition, a higher reproduction had consequences for the percentage surviving aphids, especially when one realises that even 137 Chapter9 infected aphids give birth (Hall, 1976).Furthermore , amor e favourable host plant does not encourage mobility of theinsec t andtherefor e pick upo f conidia (Hall& Papierok , 1982). Plant species differ substantially in their chemical composition, among which the nature and quantity of toxic substances in their phloem sap (Harborne et al., 1999). For instance,som etomat ocultivar sproduc e tomatine which wasfoun d tob etoxi c toBeauveria bassiana (Costa& Gaugler , 1989)an dinhibite d germination ofP. fumosoroseus a thig hlevel s (Poprawski et al., 2000). Also leaves of poinsettia contain cytotoxic triterpenoids (Smith- Kielland et al., 1996) and extracts of these leaves were used against fungi and nematodes (Khan et al., 1996; Sebastian & Gupta, 1996). These substances may also affect the susceptibility of apes t organism topathogen s (Hare& Andreadis , 1983;Ramosk a &Todd , 1985).Fruits ,root s orseed s of theCucurbitacea e contain cucurbitacins which werefoun d to protect thespotte dcucumbe rbeetl eagains tM. anisopliae (Tallamy etal., 1998),bu t weuse d ales sbitte rcucumbe rcultiva rwhic h containsver ylo w amounts ofcucurbitacin s (A.Janssen , pers.comm.) .Bot hB. bassiana andP. fumosoroseus performed better against B. argentifolii oncucumbe rtha n ontomat oplant s (Poprawski etal., 2000).B. argentifolii nymph sreare d on poinsettia areunsuitabl e asfoo d for lacewings, whichha sbee n attributed to accumulation of detrimental plant compounds in the whitefly (Legaspi etal., 1996).Wit h the application of entomopathogenic fungi againstwhitefl y onpoinsetti a asimila r interaction between first and third trophic level may exist, hence explaining the low whitefly mortality on thishos t plant (chapter 5- 7) . Another wayth ehos tplan tinfluence s theefficac y ofentomopathogeni c fungi isb yit s leaf structure.Thi swil linfluenc e thewa yspra ydroplets ,an dwit hthe mth efunga l spores,ar e deposited.Th eamoun t of trichomes andth ecompositio n of epicuticular waxes can influence thehydrophobicit y of aleaf .Venation , epidermis cells and ornamentation of alea f enhances theroughnes s ofa lea f andwit hi tth ehydrophobicit y (Holloway, 1970).Fo r instance hirsute leaves produce a more repellent surface than smooth leaves (Holloway, 1970) and droplet retention increases with decreasing surface tension (Brunskill, 1956). On relatively smooth leavesspra ydroplet stende dt ob elarge ran dmerge deasier , thano nhirsut eleave s(chapte r8) . Inaddition , nymphs atth elea f plane are shielded by the leaf trichomes andles s conidia will reach them. Although poinsettia is relatively smooth compared to leaves of certain gerbera cultivars,i tha dth emos thydrophobi clea f surface.Wate rdroplet so nthi slea f stayedcompact , even whena spreade r wasused ,i ncontras t todroplet so ngerbera ,cucumbe ro rtomat o leaves (chapter 6). This will lead to a less uniform distribution of conidia on the poinsettia leaf surfacecompare dt oleave so fth eothe rplan tspecies .A mor eunifor m distributionwil lprovid e greateropportunit yfo r insectst ocom e intocontac t with conidia on the leaf surface (Ebertet al., 1999) during thecrawle r stage or duringmoulting . 138 General discussion Humidity Oneo fth emajo r limitationso fentomopathogeni cfung i isthei rrequiremen tfo rhig hhumidit y for thegerminatio n andinfectio n processes.Th ephyllospher e environment ofth efungus , and especiallyth erelativ ehumidity ,i so f paramount importance to itseffectivit y and persistence (Diem, 1971; Drummond et al., 1987). Ambient relative humidity, temperature and air turbulencewil linfluenc e theboundar ylaye ri nwhic h theinsect-fungu s interaction takesplac e (Burrage, 1971;Ferr o& Southwick , 1984).However , theinfluenc e ofhumidit yo n successo f entomopathogenic fungi is intertwined with host-plant characteristics, since phyllosphere humidity and thickness of thelea f boundary layer are also affected by crop-specific features such as size, shape and position of leaves, leaf surface topography (hairs, waxes, stomata, veins),lea f areainde x andphotosyntheti c activity(transpiration ) (Ferro& Southwick , 1984). In general ahighe r ambient humidity will lead to ahighe r mortality duet o infection byentomopathogeni cfung i (chapter7) .However , toreac ha certai nleve lo fwhitefl y mortality the duration and the level of high humidity differs per host-plant species. On poinsettia, mortalitylevel so fT. vaporariorum variedbetwee n 50-85%a thig h ambient humidity levels (with 48 hrs >95 %RH ;chapte r 4 &7 ) and atlo w ambient humidity levels (45%) infection levelsdi dno texcee d 25%.I ncontrast , oncucumbe r mortalitylevel sa tlo w ambient humidity were still 50%eve n without any additional period of high humidity (chapter 7). The duration and the level of high humidity to reach a certain level of whitefly mortality alsodiffer s perfunga l species.Entomopathogeni c fungi differ considerably in their humidity requirements, some fungi may need a saturated environment to germinate, like V. lecanii (Hall, 1981), while others, such as M. anisopliae (Walstad et al., 1970) and A. aleyrodis (Fransen, 1995), are able to germinate under slightly dryer conditions. For instance, under dry conditions (avg 45% RH) V. lecanii does not perform as well as A. aleyrodis evenwit h anadditiona l4 8hr s> 95 % RH,wherea s under humid conditions (avg 85% RH) V.lecanii eventuall y reaches the same level of whitefly infection asA. aleyrodis (chapter7) . Althoughman yentomopathogeni c fungi need aR Ho f atleas t 92%o r more, optimal germination andgrowt hwil ltak eplac e at 100%R H (Gillespie &Crawford , 1986;Milne r& Lutton, 1986; Fransen, 1987). However, a phyllosphere humidity of 90% on gerbera and tomato was apparently sufficient for both Aschersonia species to cause a high whitefly mortality (chapter 6),comparabl et omortalit ylevel s on cucumber (phyllopshere RH >92%) . The high whitefly mortality on gerbera and tomato suggests that the calculated humidity in chapter 6i sa nunderestimatio n ofth erea lphyllospher e humidity. One should realise that the humiditydat ashow ni nchapte r6 ar ea naverag eo fa whol elea f andthu sobscurin g differences within a leaf caused by roughness (e.g. hairiness, venation), stomata, and other leaf 139 Chapter9 characteristics (Holloway, 1970).Furthermore ,th einfluenc e ofth einsec to nth ephyllospher e climatean dth emicroclimat e of theinsec titsel f (Kramer, 1980;Ramoska , 1984)whic h could affect fungal germination and growth, areno ttake nint o account. Onth eothe rhan d forpoinsetti a the calculated phyllosphere humidity mayhav ebee n overestimated. On this crop, all three fungi caused whitefly mortality no higher than 21%, whereasbot hAschersonia species were able tocaus e high whitefly mortalities on cucumber, gerberaan dtomat ounde rsimila ro rles sfavourabl e phyllospherehumiditie s(chapte r6) .Baili e etal. (1994a; 1994b)foun d thatth etranspiratio n ratefo rpoinsetti ai slo wcompare d withothe r ornamentals,an dlea f stomatalresistanc e wasconsiderabl yhighe rfo r poinsettia (highertha n rs = 100 s/m, as was the assumption in the vapour pressure deficit formula, see Appendix chapter6) .Thi sresult si na lowe rhumidit yleve ltha nwa scalculated ,whic hma yinhibi t fungal growtho rma yeve nb edetrimenta lt oth efungus . Inthi srespec t thehydrophobicit yo fth eleaf , whichi sinfluence d byit swa xlaye ran dhairiness ,i sals o likelyt opla ya considerabl e rolei n the actual phyllosphere humidity. Of all plant species used here, poinsettia had the most hydrophobic leaf surface. Water droplets on this leaf stayed compact, even when a spreader wasused ,i ncontras t todroplet s on gerbera, cucumber or tomato leaves.Th e hydrophobicity ofpoinsetti a leavestogethe r witha (possible )highe rstomata l resistance and 'sunken' stomata mightindicat etha tpoinsetti ai sadapte d toa relativel y dryenvironmen t byreducin g its water lossb ytranspiratio n asmuc h aspossible .Thi smigh t also affect germination and infection of whitefly byentomopathogeni c fungi. The phyllosphere humidity on cucumber (RH > 93%) should be high enough for conidia ofA .aleyrodis t o germinate on the leaf surface (chapter 6),bu t this did not happen (Chapter 5).Als oo ngerber a andpoinsetti a leaf conidiao fAschersonia spp.di dno t germinate (chapter 4 and 5), indicating that they need additional triggers for germination, since they germinated readilyo nnymph so fB. argentifolii (chapter 4). Although phyllosphere humidity may not explain the differences in mortality between host-plant species in its own, it does explain differences within host-plant species and between fungal species. Commercial application Mycopesticidesagains twhitefl yhav ebee nintroduce di nsevera lways .A tth ebeginnin go fthi s century, when spray application was not yet successful (because sprayequipmen t contained traceso finsecticide s andfungicide s ofearlie r applications),transplantin g ofyoun gcitru stree s with infected whitefly nymphs into citrus plantations not yet infested byA. aleyrodiswa s recommended.Anothe r wayt ointroduc eth efungu s waspinnin gleave swit hinfecte d nymphs into trees infested with D. citri(Fawcett , 1908). 140 General discussion Later, spray applications wererecommende d and care wastake n tohi tth e underside ofth eleaves ,wher ewhitefl y nymphsar emos tcommon . Conidial suspensions weremad eb y mashing up the cultures in water and sieving them through cheese cloth. One culture bottle provided sufficient spores to treat 70-100 citrus trees against D. citri.I n citrus, application from Aprilt oOctobe rwer epossibl e(Florida) ,bu tth eeffec t wasmos tcertai ndurin gth erain y summerperio d from July to mid-September (Berger, 1920b; 1921).Th erecommende d dose inglasshouse s is 1013conidia/h a(10 6- 108conidia/ml )t oreac h an almost complete kill ofT. vaporariorum in a cucumber crop (Kogan & Seryapin, 1978;Ramaker s & Samson, 1984). Control is little affected by outdoor conditions, because the naturally occurring relative humidity in a full grown cucumber crop is sufficiently high for infection (Ramakers et al., 1982; Ramakers, 1983). Also in a gerbera crop, application of 1012- 1013conidia/h a was sufficient toreac h 80%mortalit y ofT. vaporariorum (Chapter 8). At the moment several fungal products against whitefly are on the market. These products are based on, for instance, B. bassiana (Naturalis, Botanigard), P.fumosoroseus (PreFeRal orPFR )an dV. lecanii (Mycotal).Onl yMycota l isregistere d onth eDutc h market, whereas other products are registered abroad for the control of whiteflies, but are not yet available to Dutch growers. Results with Mycotal have been variable, depending on crop species andhumidit ylevel si nth e glasshouse (Ravensberg etal., 1990) .Th ebes t results were obtainedwhe nhumidit ylevel swer ea tleas t severalhour sabov e 80%.A tlowe rhumiditie sth e performance canb eimprove db yaddin ga vegetabl eoi l(Addit) .Aschersonia spp.hav eshow n tob ehighl yeffectiv e inkillin gwhiteflies , more sotha nV. lecanii(Twee n formulated), more consistently and at lower humidity levels (chapters 6 - 8).B. bassianaan d P.fumosoroseus seem to have the same humidity requirements asAschersonia spp. (Wraight et al., 2000). However, since the interaction between host plant and ambient humidity has a significant impacto nth eobtaine dwhitefl y mortality (chapter 7),on e shouldb ecarefu l comparing fungi tested in different bioassays and on different hostplants . The experiments of Wraight et al. (2000), for instance, were carried out on hibiscus, whereas in this thesis experiments were conducted on cucumber, gerbera and poinsettia.Aschersonia spp. seem to prefer a 'sit-and- wait' strategy. Hatching and moulting nymphs will settle on them. This strategy enables Aschersonia spp.t o survive and cause mortality levels up to 90% 31 days after application (chapter 5).I ncontrast ,V. lecaniigerminate so nth elea f surface (Verhaar, 1998), whichcoul d compromise its persistence, since germlings are more susceptible to adverse circumstances than ungerminatedconidi a (Sussman, 1968).It shalf-lif e is considerably shorter than thato f A. aleyrodis, viz. 4 days on chrysanthemum (Gardner et al., 1984). B. bassiana generally provides control for 3-4 days after application (Vandenberg etal., 1998), although in some cases 10% from the original dose was still present in the field after 28 days (James et al., 141 Chapter 9 Table 9.1: continued. Aspects for evaluation Results Reference Environmental characteristics: • Biotic factors host plant • host-plant chemistry • not tested, but likely influences • chapter 8 whitefly development, and maybe fungal infection chapter 5-7 crop structure, • influence phyllosphere humidity - chapter 6 leaf size and shape, • no correlation between crop architecture - chapter 6 trichomes density, and phyllosphere humidity between host- cuticular waxes plant species cultivar effects on mycosis, dependent - chapter 8 on fungal species microbial interactions not tested 144 References Abbott, W. S. (1925). A method of computing the effectiveness of an insecticide. Journal of Economic Entomology 18,265-267 . Adam,H . (1978).Moglichkeite n zurbiologisch e Bekampfung der Weisse Fliege und Roten Spinne unter Glas undPlaste n -Erfahrunge n inde r Sowjetunion und anderen sozialistischen Landern. Gartenbau 25,44-45. Agudelo,F . &Falcon ,L .A . (1977).Som e naturally occurring insect pathogens inColombia . Turrialba 27,423 - 424. Aleshina, O. A. (1978). Composition and prospects for study of the entomopathogenic fungi of the USSR. Mikologiya iFitopatologiya 12, 257-260. Allen,P .J . (1976).Contro l of spore germination and infection structure formation inth e fungi. In Physiological Plant Pathology (P.M. William & R. Heitefuss, Eds),pp . 51-128.Springe r Verlag, New York, USA. Anonymous (1999).Gewasbeschermingsgids: handboek voor de bestrijding van ziekten, plagen en onkruiden ende toepassing van groeiregulatoren inde akkerbouw, veehouderij, tuinbouw en het openbaar groen. Plant Protection Service, Wageningen, NL. Askary, H., Benhamou, N. & Brodeur, J. (1999). Ultrastructural and cytochemical characterization of aphid invasion byth ehyphomycet e Verticilliurnlecanii .Journal ofInvertebrate Pathology 74, 1-13. Asselberg, D. J. M., Nierop, S. van, Oomen, P. A. & Oostelbos, P. F. J. (1996). Gewasbeschermingsgids. Plant Protection Service, Wageningen, NL. Ayliffe, M. A., Lawrence, G. J., Ellis,J . G. & Pryor, A. J. (1997). Production of self-inhibitor of urediospore germination in Melampsora ftm'(flax rust) segregates as a recessive, single gene trait. Canadian Journal of Botany 75, 74-76. Bailie, M., Bailie, A. & Laury, J. C. (1994a). Canopy surface resistances to water vapour transfer for nine greenhouse pot plant crops. Scientia Horticulturae 57, 143-155. Bailie,M. ,Bailie ,A . &Laury , J. C. (1994b). A simplified model for predicting evapotranspiration rate of nine ornamental species vs.climat e factors and leaf area. Scientia Horticulturae 59, 217-232. Baker, E. A. & Jeffree, C. E. (1981). Characterizations of extraneous particles located on the leaf surface of several plant species.New Phytologist 87, 115-119. Barbosa, P., Saunders, J.A. & Waldvogel, M. (1982). Plant-mediated variation in herbivore suitability and parasitoid fitness. InProceedings of the5th International Symposium onInsect-Plant Relationships (J.H. Visser & A.K. Minks, eds.), pp. 63-71. Pudoc, Wageningen, the Netherlands, Wageningen, the Netherlands. Bartlett, M. C. & Jaronski, S. T. (1988). Mass production of entomogenous fungi for biological control of insects.I nFungi inbiological control systems (M.N.Burge ,Ed) ,pp .61-85 .Mancheste r University Press, Manchester, UK. Bateman,R . P.,Carey ,M. , Moore, D. &Prior , C (1993). The enhanced infectivity of Metarhiziumflavoviride in oil formulations to desert locusts atlo w humidities.Annals ofApplied Biology 122, 145-152. Becker, H.,Corliss ,J. , De-Quattro, J., Gerrietts,M. , Senft, D., Stanley, D. &Wood , M. (1992).Ge t the whitefly swatters -fast . Agricultural Research Washington 40, 4-13. Beckett, A. & Read, N. D. (1986). Low temperature scanning electron micorscopy. In Ultrastructural Techniquesfor Microorganisms (W.C .Aldric h &W . J.Todd , Eds),pp . 45-86. PlenumPublishing , New York. Bellows Jr., T.S., Perring, T.M., Gill, RJ. & Headrick, D.H. (1994). Description of a species of Bemisia (Homoptera: Aleyrodidae).Annals of the Entomological Society ofAmerica 87, 195-206. Ben-Ze'ev,I . S., Gindin,G. , Barash, I.& Raccah,B (1994).Identificatio n of entomopathogenic fungi attacking Bemisia tabaci in Israel.Bemisia Newsletter (Special Issue) 8, 36. Ben-Ze'ev, I. S., Zelig, Y., Bitton, S. & Kenneth, R. G. (1988). The Entomophthorales of Israel and their arthropod hosts: additions 1980-1988.Phytoparasitica 16,247-257 . 145 References Berenbaum, M. R. (1988). Allelochemicals in insect-microbe-plant interactions; agent provocateurs in the coevolutionaryarm s race.I nNovel Aspects ofInsect-Plant Interactions (P. Barbosa & D.K. Letourneau, Eds), pp. 97-123. John Wiley & Sons, New York, USA. Berger, E. W. (1920a). Whitefly fungus. Quarterly Bulletin of the Florida State Board 4, 104. Berger, E. W. (1920b). Work of the entomological department state plant board. Florida State Horticultural Society Quarterly, 160-170. Berger, E.W . (1921).Natura l enemies of scale insects and whiteflies. Quarterly Bulletin of Florida State Plant BoardS, 141-154. Blackman, R. L.& Eastop,V . F.(1984) .Aphids on the World's Crops: anIdentification Guide. John Wiley & Sons, Chichester, UK. Blakeman,J . P. (1971).Th e chemical environment of thelea f surface in relation to growth of pathogenic fungi. InEcology of leafsurface micro-organisms (T.F .Preec e &C .H .Dickinson , Eds),pp .255-268 .Academi c Press,London , UK. Boisclair, J., Brueren, G. J. & Lenteren, J. C. van (1990). Can Bemisia tabaci be controlled with Encarsia formosal Bulletin IOBC wprs 13/5, 32-35. Bolckmans, K., Sterk,G. , Eyal,J. , Sels,B .& Stepman,W . (1995).PreFeRa l (Paecilomycesfumosoroseus strain Apopka 97), a new microbial insecticide for the biological control of whiteflies in greenhouses. Mededelingen van de Faculteit Landbouwwetenschappen Rijksuniversiteit Gent 60, 707-711. Borisov,B .A . &Vinokurova , T. P.(1983) .Increasin g the effectiveness of entomopathogenic fungi. Zashchita Rastenii, (no.9 ) 20-22. Borner, C. (1956).Aleyrodida e (Aleurodina). In Handbuch der Pflanzenkrankheiten, Tierische Schadlinge an Nutzpflantzen, Teil2, Lieferung 4: Heteroptera und Homoptera (C. P. Sorauer, Ed),pp . 331-359. Paul Parey, Berlin, Hamburg. Boucias, D. G. (1998). Molecular analysis of the structure and function of fungal toxins. In VI Siconbiol: Simposio de Controle Biologico, pp. 73-76,Ri o deJaneiro , Brasil. Boucias, D. G. &Pendland , J.C (1991).Attachmen t of mycopathogens tocuticle : The initial event of mycoses in anrthropod hosts.I n TheFungal Spore and Disease Initiation in Plants and Animals (G. T. Cole & H. C. Hoch, Eds), pp. 101-180. Plenum, New York. Boucias,D . G., Bradford, D.L .& Barfield, C.S . (1984). Susceptibility of thevelvetbea n caterpillar and soybean looper (Lepidoptera: Noctuidae) toNomuraea rileyi:Effect s of pathotype, dosage, temperature, and host age. Journal of Economic Entomology 77,247-253 . Briese, D. T (1986). Host resistance to microbial control agents. In Biological Plant and Health Protection: Biological Controlof Plant Pests and of Vectors ofHuman andAnimal Diseases (J.M .Franz , Ed), Vol. 32,pp . 233-256. Fischer Verlag, Stuttgart. Brown, G. C, Prochaska, G. L.,Hildebrand , D.F. ,Nordin , G. L. &Jackson , D. M. (1995). Green leaf volatiles inhibit conidial germination of the entomopafhogen Pandora neoaphidis (Entomophthorales: Entomophthoraceae). Environmental Entomology 24, 1637-1643. Brown, J. K., Bird, J., Frohlich, D.R. , Rosell,R . S., Bedford, I.D . & Markham, P. G. (1996). The relevance of variability withinth eBemisia tabaci speciescomple x toepidemic s caused by subgroup III geminiviruses. InBemisia 1995:Taxonomy, Biology, Damage Controland Management (D.Gerlin g& R.T .Mayer ,Eds) , pp. 77-89.Intercep t Ltd., Andover, UK. Brunskill, R. T. (1956). Physical factors affecting the retention of spray droplets on leaf surfaces. In Third British Weed Control Conference. British Crop Protection Council, Brighton, UK. Burrage, S.W . (1971).Th emicro-climat ea tth elea f surface. InEcology ofLeaf SurfaceMicro-Organisms (T.F. Preece & C.H. Dickinson, Eds),pp . 91-101. Academic Press, London, UK. Butt,T .M .(1987) .A fluorescenc e microscopymetho d for therapi d localization of fungal spores and penetration sites of insect cuticle.Journal of Invertebrate Pathology 50,12-1A. Butt, T. M. (1990).Funga l infection process -a mini-review. InProceedings and Abstracts, Vth International Colloquium onInvertebrate Pathology andMicrobial Control, pp. 121-124,Adelaide , Australia. Butt,T . M.,Barrisever , M., Drummond,J. , Schuler, T.H. , Tillemans,F .T . &Wilding , N. (1992). Pathogenicity of the entomogenous, Hyphomycete fungus, Metarhizium anisopliae, against the Chrysomelid beetles Psylliodes chrysocephala and Phaedon cochleariae. Biocontrol Science and Technology 2, 327-334. 146 References Butt, T. M. & Goettel, M. S. (2000). Bioassays of entomogenous fungi. In Bioassays of entomopathogenic microbes and nematodes (A. Navon & K. R. S. Ascher, Eds), pp. 141-191. CAB International, Wallingford, UK. Butt,T .M. ,Ibrahim ,L. ,Ball ,B .V .& Clark , S.J . (1994).Pathogenicit y of theentomogenou s fungi Metarhizium anisopliae and Beauveria bassiana against crucifer pest and the honey bee. Biocontrol Science and Technology 4,207-214. Butt, T. M., Ibrahim, L., Clark, S. J. & Beckett, A. (1995). The germination behaviour of Metarhizium anisopliae on the surface of aphid and flea beetle cuticles.Mycological Research 99, 945-950. Byrne, D. N. & Bellows Jr., T. S. (1991). Whitefly biology.Annual Review of Entomology 36,431-457 . Byrne,D .N. ,Bellow s Jr.,T . S.& Parrella,M .P .(1990a) .Whiteflie s inagricultura l systems.I n Whiteflies:Their bionomics, pest status and management (D.Gerling , Ed),pp . 227-261. Intercept Ltd., Andover, UK. Byrne,D .N. ,Cohen , A.C . &Draeger , E.A . (1990b). Water uptake from plant tissue by the egg pedicel of the greenhouse whitefly, Trialeurodes vaporariorum. Canadian Journal of Zoology 68, 1193-1195. Byrne, D. N. & Hadley, N. F. (1988). Particulate surface waxes of whiteflies: morphology, composition and waxing behaviour. Physiological Entomology 13,267-276 . Carvalho, M. B. de, Aquino, M. de L. N. & Oliveira, M. H. C. C. de (1975). Considerations on the biological control of Aleurodicus cocois (Curtis, 1846) (Homoptera, Aleyrodidae), cashew whitefly. Anais do Instituto de Ciencias Biologicas 2, 25-30. Cassino,P .C .R. , Lima,E .R . & Simoes,N .M . (1984).Aleyrodid s oncitru s plants in the Serrana region of Rio de Janeiro State. InIX Congresso Brasileiro de Entomologia, Londrina - Pr. Chandler, D.,Heale , J. B. & Gillespie, A. T. (1994). Effect of osmotic potential on the germination of conidia and colony growth of Verticillium lecanii. Mycological Research 98, 384-388. Chen, X. F., Jin, J. Z., Wu, G. Y. & Sun, J. D. (1994). Bionomics of Aegerita webberi Fawcett and its application in the control of the citrus spiny whitefly. China Tea 16,4-5 . Chen, Z. A. (1988).Som e epidemilogical characteristics of entomogenous fungi Aschersonia duplex Berk, and its application. Acta Phytophylactica Sinica 15, 235-239. Chien, C. C. & Chiu, S. C. (1986). Biology of the citrus spiny blackfly Aleurocanthus spiniferus. Journal of Agricultural Research of China 35,222-229 . Clancy, K. M.& Price,P .W . (1987).Rapi d herbivore growth enhances enemy attack: sublethal plant defences remain a paradox. Ecology 68,733-736 . Clerk, G. C. & Madelin, M. F. (1965). The longevity of conidia of three insect-parasitizing hyphomycetes. Transactions of the British Mycological Society 48, 193-209. Cook, R. J., Bruckart, W. L., Coulson, J. R., Goettel, M. S., Humber, R. A., Lumsden, R. D.,Maddox , J. V., McManus,M .L. ,Moore ,L. , Meyer, S. F., Quimby Jr, P. C, Stack, J. P. & Vaughn, J. L. (1996). Safety of microorganisms intended for pest and plant disease control: A framework for scientific evaluation. Biological Control 7, 333-351. Cooke, R. C. & Rayner, A. D.M (1984). Ecology of Saprotrophic Fungi. Longman, London, UK. Costa,S .D .& Gaugler, R. (1989).Sensitivit y of Beauveria bassiana to solanine and tomatine: plant defensive chemicals inhibit an insect pathogen. Journal of Chemical Ecology 15, 697-706. De Barro,P .J . &Hart , P. J. (2000). Mating interactions between two biotypes of the whitefly, Bemisia tabaci (Hemiptera:Aleyrodidae ) in Australia.Bulletin of Entomological Research 90, 103-112. Dicke,M . & Sabelis,M .W . (1988).Ho wplant s obtain predatory mites as bodyguards. Netherlands Journal of Zoology 38, 148-165. Dicke, M., Sabelis, M. W., Takabayashi, J., Bruin, J. & Posthumus, M. A. (1990). Plant strategies of manipulating predator-prey interactions through allelochemicals: prospects for application in pest control. Journal of Chemical Ecology 16, 3091-3118. Diem, H. G. (1971). Effect of low humidity on the survival of germinated spores commonly found in the phyllosphere. In Ecology of Leaf Surface Micro-Organisms (T.F. Preece & C.H. Dickinson, Eds), pp. 211-219. Academic Press, London, UK. Dik, A.J . &Elad ,Y . (1999).Compariso n of antagonists of Botrytiscinerea i n greenhouse-grown cucumber and tomato under different climatic conditions.European Journal of Plant Pathology 105, 123-137. 147 References Dingley, J. M. (1954). The Hypoceales of New Zealand. VI. The Genera Hypocrella, Barya, Claviceps and Podonectria. Transactions of the Royal Society ofNew Zealand 81, 489-499. Dittrich, V. & Ernst, G. H. (1990). Chemical control and insecticides of whiteflies. In Whiteflies: Their bionomics, pest status and management (D. Gerling, Ed), pp. 263-285. Intercept Ltd., Andover, UK. Dixon, A.F .G . (1987).Th e wayo f life of aphids: Host specificity, speciation anddistribution . InAphids. Their biology,Natural Enemiesand Control(A.K . Minks &P .Harrewijn , Eds),Vol . 2A,pp . 197-208.Elsevier , Amsterdam. Dorsman, R. & Vrie, M. van de (1987). Population dynamics of the greenhouse whitefly Trialeurodes vaporariorum on different Gerbera varieties.Bulletin IOBC wprs 10/2, 46-5. Dowell, R. V. (1982).Biolog y of Tetraleurodes acaciae (Quaintance) (Homoptera: Aleyrodidae). Pan Pacific Entomologist 58, 312-318. Drost,Y .C , Lenteren,J . C.va n &Roermund , H.J . W. van (1998).Life-histor y parameterso f different biotypes of Bemisia tabaci (Hemiptera:Aleyrodidae ) inrelatio n totemperatur e and host plant: A selective review. Bulletin of Entomological Research 88, 219-229. Drummond,J. ,Heale ,J . B.& Gillespie,A .T . (1987).Germinatio n andeffec t ofreduce d humidityo n expression ofpathogenicit y in Verticillium lecaniiagains tth e glasshouse whitefly Trialeurodesvaporariorum. Annals ofApplied Biology 111, 193-201. Duffus, J. E. (1996). Whitefly-borne viruses. In Bemisia 1995: Taxonomy, Biology, Damage Control and Management (D. Gerling & R.T. Mayer, Eds),pp . 255-263.Intercep t Ltd., London, UK. Ebert, T. A., Taylor, R. A. J., Downer, R. A. & Hall, F. R. (1999).Deposi t structure and efficacy of pesticide application. 1:Interaction s between deposit size, toxicant concentration and deposit number. Pesticide Science 55,783-792 . Eijk, G. W. van, Mummery, R. S., Roeymans, H. J. & Valadon, L. R. G. (1979). A comperative study of carotenoids ofAschersonia aleyrodisand Aspergillus giganteus. Antonie van Leeuwenhoek: Journal of Microbiology 45,417-422 . Ekbom, B.S . (1979).Biologica l control of the glasshouse whitefly (Trialeurodes vaporariorum). Brief history and future prospects.Entomologisk Tidskrift 100, 1-10. Ekbom, B. S. & Ahman, I. (1980). The fungus Verticillium fusisporum as an insect pathogen. Journal of Invertebrate Pathology 36, 136-138. Elad, Y., Malathrakis, N. E. & Dik, A. J. (1996). Biological control of Borryrw-inciteddisease s and powdery mildews in greenhouse crops. Crop Protection 15, 229-240. Elizondo, J. M & Quezada, J. R. (1990). Identification and evaluation of the natural enemies of the citrus blackfly Aleurocanthus woglumi Ashby (Homoptera: Aleyrodidae) in four citrus-growing areas of Costa Rica. Turrialba 40, 190-197. Elliot, S.L. , Sabelis,M .W. ,Janssen , A., Geest, L. P. S. van der, Beerling, E. A. M. &Fransen , J. (2000). Can plants use entomopathogens as bodyguards? Ecology Letters 3,228-235 . Entcheva, L. (1979). The influence of some factors affecting the development of the superparasitic fungus Aphanocladium album Gams, as a bioagent against larvae of the glasshouse whitefly. Gradinarska i Lozarska Nauka 16,78-82 . Evans, H. C. (1988). Coevolution of entomogenous fungi and their insect hosts. In Coevolution offungi with plants and animals (K.A. Pirozynski & D.L. Hawksworth, Eds),pp . 149-171.Academi c Press, London. Evans,H .C . (1994).Spor egeminatio n in theentomopathogeni c genus Aschersonia. Mycological Research 98, 165-168. Evans, H. C. & Hywel-Jones, N. (1990). Aspects of the genera Hypocrella and Aschersonia as pathogens of coccids and whiteflies. In Proceedings and Abstracts, Fifth International Colloquium on Invertebrate Pathology and Microbial Control, pp. 111-115. Evans, H. C. & Hywel-Jones, N. L. (1997). Entomopathogenic fungi. In Soft Scale Insects. Their Biology, Natural Enemies and Control (Y. Ben-Dov & C.J. Hodgson, Eds), Vol. 7B, pp. 3-27. Elsevier Science, Amsterdam. Ewald,P .W . (1995).Th eevolutio n of virulence: a unifying linkbetwee n parasitology and ecology.Journal of Parasitology 81, 659-669. 148 References Fang, Q.X. , Hu, T. C, Gong, Y.X. , Zhou, Y. K., Hu, Y.M. , Yang, S.F. , Zhou, Y. Y., Wang, H. M., Zhou, G. G., Bao, L. Y. & Guan, X. M. (1983). Studies on the control of the greenhouse whitefly Trialeurodes vaporariorum withAschersonia. Acta Entomologica Sinica 26, 278-286. Fargues, J. (1976). Specificite des champignons pathogenes imparfaits (hyphomycetes) pour les larves de coleopteres (Scarabaeidae et Chrysomelidae). Entomophaga 21,313-323 . Fargues, J. & Remaudiere, G. (1977). Considerations on the specificity of entomopathogenic fungi. Mycopathologia 62, 31-37. Fargues,J . &Rodriguez-Rueda , D. (1980).Sensibilit e deslarve s deSpodoptera littoralis [Lep.:Noctuidae ] aux Hyphomycetesentomopathogene sNomuraea rileyie tPaecilomyces fumoso-roseus. Entomophaga 25,43 - 54. Fargues, J., Rougier, M., Goujet, R. & Itier, B. (1988). Effect du rayonnement solaire sur la persistance des conidiaspores de l'hyphomycete entomopathogene, Nomuraea rileyi, a la surface couvert vegetal. Entomophaga 33, 357-370. Fargues,J. , Maniania, N.K. ,Delmas , J. C.& Smiths,N . (1992).Influenc e of temperature onth ei n vitro growth of entomopathogenic hyphomycetes.Agronomie 12, 557-564. Fargues, J., Goettel, M. S., Smits, N., Ouedraogo, A., Vidal, C, Lacey, L. A., Lomer, C. J. & Rougier, M. (1996).Variabilit y in susceptibility tosimulate d sunlight of conidia among isolates of entomopathogenic Hyphomycetes.Mycopathologia 135, 171-181. Fawcett, H. S. (1908).Fungi parasiti c uponAleyrodis citri. University of Florida Experimental Studies 1, 1-41. Fawcett, H. S. (1944).Fungu s andbacteria l diseases of insects asfactor s inbiologica l control.Botanical Review 10, 327-348. Fedorova, E. M. (1990). Technical effectiveness of different groups of preparations against citrus whitefly. Subtropicheskie Kul'tury, 117-120. Feng,M . G., Poprawski,T . J. & Khachatourians, G. G. (1994). Production, formulation and application of the entomopathogenic fungus Beauveria bassiana for insect control: current status.Biocontrol Science and Technology 4, 3-34. Ferro, D.N . &Southwick ,E .E .(1984) .Microclimate s of small arthropods: estimating humidity within the leaf boundary layer. Environmental Entomology 13,926-929 . Ferro, D. N., Chapman, R. B. & Penman, D. R. (1979). Observations on insect microclimate and insect pest management. Environmental Entomology 8, 1000-1003. Ferron, P. (1977).Influenc e of relative humidity on thedevelopmen t of fungal infection caused by Beauveria bassiana [Fungi Imperfecti, Moniliales] in imagines of Acanthoscelides obtectus [Col.: Bruchidae]. Entomophaga 22, 393-396. Ferron, P.(1985) .Funga lcontrol .I n ComprehensiveInsect Physiology, Biochemistry and Pharmacology (G.A. Kerkut & L.I. Gilbert, Eds), Vol. 12,pp . 313-346. Pergamon Press, Oxford. Ferron,P. ,Fargues ,J . & Riba, G. (1991).Fung i as microbial insecticides. In Handbook ofApplied Mycology, Vol2: Humans, Animals, andInsects (D.K . Arora, L.Ajell o &K .G . Mukerji, Eds),pp .665-706 . Marcel Dekker Inc.,Ne w York. Ferron,P. ,Hurpin , B.& Robert, P.H . (1972). Surla specificite' de Metarrhizium anisopliae (Metsch.) Sorokin. Entomophaga 17, 165-178. Foschi, S. &Deseo ,K . V. (1987).Prospect so f microbiological control of phytophagous pests in flowering and ornamental crops in a protected environment. Difesa delle Piante 10, 121-125. Fransen, J. J. (1987). Aschersonia aleyrodis as a microbial agent of greenhouse whitefly, Wageningen Agricultural University, Wageningen, The Netherlands. Fransen, J. I. (1990). Natural enemies of whitefly: fungi. In Whiteflies: their bionomics, pest status and management (D. Gerling,Ed) ,pp . 187-210. Intercept, Andover, UK. Fransen, J. J. (1993).Integrate d pest management in glasshouse ornamentals in the Netherlands: a step by step policy. Bulletin IOBC wprs 16/2,35-38 . Fransen, J. J. (1994).Bemisia tabaci in the Netherlands; Here to stay? Pesticide Science 42,129-134. Fransen, J. J. (1995). Survival of spores of the entomopathogenic fungus Aschersonia aleyrodis (Deuteromycotina: Coelomycetes) on leaf surfaces. Journal of Invertebrate Pathology 65,73-75 . 149 References Fransen, J. J. & Lenteren, J. C. van (1993). Host selection and survival of the parasitoid Encarsiaformosa on greenhouse whitefly, Trialeurodes vaporariorum, in the presence of host infected with the fungus Aschersonia aleyrodis. Entomologia Experimentalis etApplicata 69,239-249 . Fransen, J. J. & Lenteren, J. C. van (1994). Survival of the parasitoid Encarsia formosa after treatment of parasitized greenhouse whitefly larvae with fungal spores of Aschersonia aleyrodis. Entomologia Experimentalis etApplicata 71, 235-243. Fransen, J. J. & Montfort, M. A. J. van (1987). Functional response and host preference of Encarsia formosa Gahan (Hym., Aphelinidae), a parasitoid of greenhouse whitefly T. vaporariorum (Westwood) (Horn., Aleyrodidae).Journal ofApplied Entomology 103,55-69 . Fransen, J. J., Winkelman, K. &Lenteren , J. C. van (1987). The different mortality at various life stages of the greenhouse whitefly, Trialeurodes vaporariorum (Hymenoptera: Aleyrodidae), by infection with the fungus Aschersonia aleyrodis (Deuteromycotina: Coelomycetes). Journal of Invertebrate Pathology 50, 158-165. Furlong, M. J. & Pell, J. K. (1996). Interactions between the fungal entomopathogen Zoophthora radicans Brefeld (Entomophthorales) and twohymenoptera n parasitoids attacking the diamondback moth, Plutella xylostella. Journal of Invertebrate Pathology 68, 15-21. Furlong, M. J. & Pell, J. K. (1997). The influence of environmental factors on the persistence of Zoophthora radicans conidia. Journal of Invertebrate Pathology 69,223-233 . Fuxa, J. R. (1987). Ecological considerations for the use of entomopathogens in IPM. Annual Review of Entomology 32, 225-251. Gaede, K. (1992). On the water balance of Phytoseiulus persimilis A.-H. and its ecological significance. Experimental andApplied Acarology 15, 181-198. Gallardo, F.,Boethel ,D .J. , Fuxa, J. R. & Richter, A. (1990). Susceptibility of Heliothis zea (Boddie) larvae to Nomuraea rileyi (Farlow) Samson: Effects of a-tomatine at the third trophic level.Journal of Chemical Ecology 16, 1751-1759. Gao, R. X. & Ouyang, Z. (1981).A preliminar y investigation on parasitic fungi on citrus pests. Weishengwuxue Tongbao 8, 57-58. Gardner, W. A., Oetting, R. D. & Storey, G. K. (1984). Scheduling of Verticillium lecanii and Benomyl applications tomaintai n aphid (Homoptera: Aphidae) control onChrysanthemu m in greenhouses. Journal of Economic Entomology 77. Gates, D. M. (1980). Biophysical Ecology. Springer, New York. Gerling, D. (1990). Natural ennemies of whiteflies: predators and parasitoids. In Whiteflies: Their Bionomics, Pest Status and Management (D.Gerling ,Ed) ,pp . 147-186.Intercep t Ltd., Andover, UK. Gerling, D (2000). Whiteflies revisited. In XXI-International Congress of Entomology, pp. LXTV-LXVIII, Brazil. Gillespie, A. T. & Claydon, N. (1989).Th e use of entomogenous fungi for pest control and the role of toxins in pathogenisis. Pesticide Science 27,203-215 . Gillespie, A. T. & Crawford, E. (1986). Effect of water activity on conidial germination and mycelial growth of Beauveria bassiana, Metarhizium anisopilae, Paecilomyces spp. and Verticillium lecanii. In Fundamental and Applied Aspects of Invertebrate Pathology, Proceedings of the Fourth International Colloquiumon Invertebrate Pathology (R.A. Samson,J.M . Vlak & D. Peters, eds.),pp . 254, Veldhoven, The Netherlands. Gindin, G. & Ben-Ze'ev, I. S. (1994). Natural occurrence of and inoculation experiments with Conidiobolus coronatusan d Conidiobolussp .i nglasshous epopulation s ofBemisia tabaci.Phytoparasitica 22 , 197-208. Gindin, G., Barash, I., Harari, N. & Raccah, B. (1994). Effect of endotoxic compounds isolated from Verticillium lecanii on the sweetpotato whitefly, Bemisia tabaci. Phytoparasitica 22, 189-196. Gindin,G. , Geschtovt,N .U. ,Raccah ,B .& Barash, I. (2000).Pathogenicit y of Verticillium lecanii to different developmental stages of the silverleaf whitefly, Bemisia argentifolii. Phytoparasitica 28, 229-239. Glare, T. R. & Chilvers, G. A. (1985). Capilliconidia as infective spores in Zoophthora phalloides (Entomophthorales). Transactions of the British Mycological Society 85,463-470 . Goettel, M. S. (1992). Fungal agents for biocontrol. In Biological control of Locusts and Grasshoppers (C.J. Lomer & C. Prior, Eds),pp . 122-132.CA B International, Wallingford, UK. 150 References Goettel, M. S., Vandenberg, J. D.,Duke ,G .M . & Schaalje, G. B.(1993) .Susceptibilit y tochalkbroo d of alfalfa leafcutter bees,Megachile rotundata, reared on natural and artificial provisions.Journal of Invertebrate Pathology 61, 58-61. Goudriaan, J. & van Laar, H.H . (1994).Modelling Potential CropGrowth Processes: Textbookwith Exercises. Kluwer, Dordrecht. Gravena, S., Leao Neto, R. R., Moretti, F. C., Tozatti, G. & Rocha Leao Neto, R. da (1988). Effectiveness of insecticides against Selenaspidus articulatus (Morgan) (Homoptera, Diaspididae) and effect on natural enemies in citrus orchard. Cientifica Jaboticabal 16,209-217 . Gullino, M. L. & Wardlow, L. R. (1999). Ornamentals. In Integrated Pest and Disease Management in Greenhouse Crops (R.Albajes , M.L . Gullino, J. C. van Lenteren & Y. Elad, Eds),pp . 486-505. Kluwer Academic Publishers, Dordrecht. Gupta,B .P . &Joshi ,N .K . (1973). Bionomics of citrus white fly, Dialeurodes citri Ashmead at Chaubattia in U.P. Hills.Progressive Horticulture 5, 77-83. Hajek, A.E . (1993).New optionsfor insect control usingfungi. American Chemical Society, Washington DC. Hajek, A. E.& Butler , L.(1999) .Predictin g thehos t rangeo f entomopathogenic fungi. InNontarget Effects of Biological Control(P .A .Folle t& J.J .Duan ,Eds) ,pp .263-276 . KluwerAcademi c Publishers, Dordrecht, The Netherlands. Hajek, A. E. & Goettel, M. S (2000). Guidelines for evaluating effects of entomopathogens on non-target organisms. In Field manual of techniques in invertebrate pathology : application and evaluation of pathogensfor control of insects and other invertebrate pests (L.A . Lacey & H. K. Kaya, Eds),pp . 847- 868. Kluwer Academic Publishers, Dordrecht, The Netherlands. Hajek, A.E . & St.-Leger, R. J. (1994). Interaction between fungal pathogens and insect hosts.Annual Review of Entomology 39, 293-322. Hajek, A. E., Renwick, J. A. & Roberts, D. W. (1995). Effects of larval host plant on the gypsy moth (Lepidoptera:Lymantriidae ) fungal pathogen,Entomophaga maimaiga (Zygomycetes: Entomophthorales). Environmental Entomology 24, 1307-1314. Hall, R. A. (1976). Verticillium lecanii on the aphid, Macrosiphoniella sanborni. Journal of Invertebrate Pathology 28, 389-391. Hall, R. A. (1980).Compariso n of laboratory infection of aphids byMetarrhizium anisopliae and Verticillium lecanii.Annals ofApplied Biology 95, 159-162. Hall, R. A. (1981). The fungus Verticillium lecanii as a microbial insecticide against aphids and scales. In Microbial Control of Pests and Plant Diseases, 1970-1980 (H.D.Burges , Ed), pp. 483-498. Academic Press, London, UK. Hall,R . A. (1982).Contro l of whitefly T. vaporariorum and cotton aphidAphis gossypii in glasshouses by two isolates of the fungus Verticillium lecanii.Annals ofApplied Biology 101, 1-12. Hall,R . A. (1985).Aphi dcontro lb yfungi . InBiological pest control,the glasshouse experience (N.W. Hussey & N. Scopes, Eds),pp . 138-141. Blandford Press, Poole, UK. Hall,R . A. &Papierok ,B .(1982) .Fung i asbiologica l control agents of arthropods of agricultural and medical importance. Parasitology 84, 205-240. Hall,R . A., Peterkin, D. & Ali, B. (1994).Funga l control of whitefly, Thripspalmi and sugercane froghopper inTrinida d and Tobago.I nProceedings ofthe Vlth International Colloquium on Invertebrate Pathology and Microbial Control, pp. 277-282, Montpellier, France. Hallsworth, J. E. & Magan, N. (1999).Wate r and temperature relations of growth of the entomogenous fungi Beauveria bassiana, Metarhizium anisopliae, and Paecilomyces farinosus. Journal of Invertebrate Pathology 74, 261-266. Harborne, J. B., Baxter, H. & Moss, G. P. (1999). Phytochemical Dictionary: a Handbook of Bioactive Compoundsfrom Plants. 2nd ed./Ed. Taylor & Francis, London. Hare,J .D . & Andreadis,T . G. (1983).Variatio n in the susceptibility of Leptinotarsa decemlineata (Coleoptera: Chrysomelidae) when reared on different host plants to the fungal pathogen Beauveria bassiana in the field and laboratory. Environmental Entomology 12, 1892-1897. Hawksworth, D. L., Kirk, P. M., Sutton, B. C, Ainsworth, G. C, Benny, G. L. & Cannon, P. F. (1995). Ainsworth &Bisby's dictionary of thefungi. - 8th ed. CAB International, Wallingford, UK. 151 References Hedge, Y. & Kolattukudy, P.E . (1997).Cuticula r waxes relieve self-inhibition of germination and appressorium formation byte hconidi ao fMagnaporthe grisea.Physiological and Molecular PlantPathology 51, 75-84. Helyer, N. (1993). Verticillium lecaniifo r control of aphids and thrips on cucumber. Bulletin IOBC wprs 16/2, 63-66. Helyer, N., Gill, G. & Bywater, A. (1992). Control of chrysanthemum pests with Verticillium lecanii. Phytoparasitica 20 suppl., 5-9. Henter, H. J & Lenteren, J.C. van (1996). Variation between laboratory populations in the performance of the parasitoid Encarsiaformosaon two species,Bemisia tabaci and Trialeurodesvaporariorum. Entomologia Experimentalis etApplicata 80, 427-434. Hirte, W. F., Glathe, I. & Adam, H. (1989a). Production und application of a microbial preparation with Aschersonia - spores.Zentralblatt fur Mikrobiologie 144, 151-162. Hirte, W. F., Walter, C, Grunberg, M., Sermann, H. & Adam, H. (1989b). Selection of pathotypes of Verticilliumlecanii for various harmful insects in glasshouses and aspects of thebiotechnologica l spore production. Zentralblatt fur Mikrobiologie 144, 405-420. Holloway, P.J . (1970). Surface factors affecting the wetting of leaves.Pesticide Science 1, 156-163. Hoover, K.,Alaniz , S. A., Yee, J. L., Rocke, D. M., Hammock, B. D. & Duffey, S. S. (1998). Dietary protein and chlorogenic acid effect on baculoviral disease of noctuid (Lepidoptera: Noctuidae) larvae. Environmental Entomology 27, 1264-1272. Horowitz, A. R. & Ishaaya, I. (1996). Chemical control of Bemisia -managemen t and application. In Bemisia 1995: Taxonomy,Biology, Damage Controland Management (D.Gerlin g & R.T. Mayer, Eds),pp .537 - 556. Intercept Ltd., Andover, UK. Houten, Y. M. van, Rijn, P. C. J. van, Tanigoshi, L. K., Stratum, P. van & Bruin, J. (1995). Preselection of predatory mites toimprov e year-round biological control of Western Flower Thrips in greenhouse crops. Entomologia Experimentalis etApplicata 74, 225-234. Hulden, L. (1986).Th e whiteflies (Homoptera, Aleyrododae).Notulae Entomologicae 66, 1-40. Humber, R. A. (1992). USDA-ARS Collection of Entomopathogenic Fungal Cultures. USDA-ARS Plant Protection Research Unit, Ithaca, New York. Humphreys, A. M., Inch, J. M. M., Gillespie, A. T. & Trinci, A. P. J. (1986). Blastospore production in Paecilomyces species. In Fundamental and appliedaspects ofinvertebrate pathology, Proceedings of the Fourth International Colloquiumon Invertebrate Pathology (R.A. Samson, J.M. Vlak & D. Peters, eds.), Veldhoven, The Netherlands. Hussey, N. W. (1958).Note s on afungu s parasitic on whitefly. Plant Pathology 7, 71-72. Hywel-Jones, N. L. (1993). A systematic survey of insect fungi from natural, torpical forest in Thailand. In Aspects of Tropical Mycology (S. Isaac, J. C. Frankland & R. Watling, Eds), pp. 300-301. Cambridge University Press, Cambridge. Hywel-Jones,N .L . & Evans,H .C . (1993).Taxonom y and ecology of Hypocrella discoidea and its anamorph, Aschersonia samoensis. Mycological Research 97, 871-876. Ibrahim, Y. B. & Lee, C. Z. (1996). Infectivity of the spiralling whitefly Aleurodicus dispersus Russell (Homoptera: Aleurodidae) byth e entomopathogenic fungus Aschersonia placenta B. and Br. Malaysian Applied Biology IS, 13-17. Ibrahim, Y. B., Lim, T. K., Tang, M. K. & Teng, H. M. (1993). Influence of temperature, pH and selected growth mediao ngermination , growth andsporulatio n ofAschersonia placenta andHypocrella raciborskii. Biocontrol Science and Technology 3, 55-61. Inbar, M., Doostdar, H., Leibee, G.L. & Mayer, R.T. (1999a). The role of plant rapidly induced responses in asymmetric interspecific interactions among insect herbivores.Journal of Chemical Ecology 25,1961 - 1979. Inbar,M. , Doostdar, H. &Mayer , R. T. (1999b).Effect s of sessile whitefly nymphs (Homoptera: Aleyrodidae) on leaf-chewing larvae (Lepidoptera :Noctuidae) . Environmental Entomology 28, 353-357. Inglis, G. D., Goettel,M . S. & Johnson, D. L. (1993). Persistence of the entomopathogenic fungus, Beauveria bassiana, on phylloplanes of crested wheatgrass and alfalfa. Biological Control 3, 258-270. Isaac, S. (1992).Fungal-plant interactions. Chapman & Hall, London, UK. 152 References James, R. R., Shaffer, B. T., Croft, B. & Lighthart, B. (1995). Field evaluation of Beauveria bassiana: its persistence and effects on the pea aphid and a non-target coccinellid in alfalfa. Biocontrol Science and Technology 5, 425-437. Jaques, R. P.(1972) .Th e inactivation of foliar deposits of viruses of Trichoplusia ni (Lepidoptera: Noctuidae) and Pieris rapae(Lepidoptera: Pieridae) and tests on protectan additives. Canadian Entomologist 104, 1985-1994. Jaques,R . P.(1983) .Th epotential ofpathogens for pestcontrol .Agriculture, Ecosystems and Environment 10, 101-126. Jones, H.G . (1986).Energ y balance and evaporation. In Plants and Microclimate: A quantitative approach to environmentalplant physiology. (H.G. Jones, Ed),pp . 85-103.Cambridg e University Press, Cambridge, UK. Jones, K.A . & Burges,H . D. (1998).Technolog y of formulation and application. In Formulation of microbial biopesticides (H.D.Burges , Ed),pp . 7-30. Kluwer Academic Publishers, Dordrecht. Kerssies, A. (1994). Epidemiology of Botrytis spotting on gerbera and rose flowers grown under glass, Wageningen Agricultural University, Wageningen. Khachatourians, G. G. (1991). Physiology and genetics of entomopathogenic fungi. In Handbook of Applied Mycology: Humans,Animals, andInsects (D.K. Arora, L. Ajello & K.G. Mukerji, Eds),Vol . 2,pp .613 - 664. Marcel Dekker Inc., New York. Khan, M. I., Ashaq, M, Bhat, M. Y. & Fazal, M. (1996). A comparative study of the efficacy of some plant extracts and fungicides on mycoflora of chickpea seed.Annals of Plant Protection Sciences 4, 85-87. Khristova, E. (1971). Possibilities of using Encasiaformosa Gah. and Aschersonia aleyrodis Webber in the control of Trialeurodes vaporariorum Westw.Acta Horticulturae 17,396-399 . Klingauf, F. A. (1987). Host plant finding and acceptance. In Aphids. Their biology. Natural Enemies and Control (A.K. Minks & P. Harrewijn, Eds),Vol . 2A, pp. 209-224. Elsevier, Amsterdam. Knauf, T. A.& Wright , J.E . (1994).Beauveria bassiana (ATCC 74040): Control of insect pests in filed crops and ornamentals. Brighton Crop Protection Conference -Pests and Diseases, 1103-1108. Knudsen, G. R., Schotzko, D. J. & Krag, C. R. (1994). Fungal entomopathogen effect on numbers and spatial patternso f the Russian wheat aphid (Homoptera: Aphididae) on preferred and nonpreferred host plants. Environmental Entomology 23, 1558-1567. Kogan, V. Sh. & Seryapin, A. A.. (1978). Against the greenhouse whitefly: Aschersonia. Zashchita Rastenii, (no. 1)29-31. Koike, S.T . & Saenz, G. S. (1998). First report of powdery mildew, caused by an Oidium sp., on poinsettia in California. Plant Disease 82, 128. Kramer, J. P. (1980).Th ehouse-fl y mycosis caused by Entomophthora muscae: influence of relative humidity on infectivity and conidial germination. Journal of theNew YorkEntomological Society 88, 236-240. Krasnoff, S. B., Gibson, D. M., Belofsky, G. N., Gloer, K. B. & Gloer, J. B. (1996). New destruxins from the entomopathogenic fungus Aschersonia sp.Journal ofNatural Products 59,485-489 . Krips, O. E. (2000). Plant effects on biological control of spider mites in the ornamental crop gerbera, Wageningen University, Wageningen, The Netherlands. Krips, O.E. ,W. , Kleijn P.,Willems ,P .E .L. ,Gols , G. J.Z . &Dicke ,M . (1999).Lea f hairs influence searching efficiency and predation rate of the predatory mite Phytoseiulus persimilis (Acari: Phytoseiidae). Experimental andApplied Acarology 23, 119-131. Kurbatskaya, Z. A., Koval', E. Z. & Fesenko, V. D. (1983). Toxin-forming capacity of the entomopathogenic fungus Aschersonia aleyrodis Webber under different conditions of culturing. Mikrobiologicheskii Zhurnal 45, 37-40. Lacey, L. A. & Mercadier, G. (1998). The effect of selected allelochemicals on germination of conidia and blastospores and mycelial growth of the entomopathogenic fungus, Paecilomyces fumosoroseus (Deuteromycotina: Hyphomycetes). Mycopathologia 142, 17-25. Lacey,L .A. ,Fransen , J.J . & Carruthers,R . (1996).Globa ldistributio n of naturally occuring fungi of Bemisia, theirbiologie san dus e asbiologica l control agents.I nBemisia 1995:Taxonomy, Biology, Damage Control and Management (D.Gerlin g & R.T.Mayer , Eds),pp . 401-433.Intercep t Ltd., Andover, UK. 153 References Lacey, L. A., Kirk, A. A., Millar, L., Mercadier, G. & Vidal, C. (1999). Ovicidal and larvicidal activity of conidia and blastospores of Paecilomyces flimosoroseus (Deuteromycotina: Hyphomycetes) against Bemisia argentifolii(Homoptera :Aleyrodidae ) with adescriptio n of abioassa y system allowing prolonged survival of control insects. Biocontrol Science and Technology 9, 9-18. Landa, Z. (1984). Protection against glasshouse whitefly (Trialeurodes vaporariorum Westw.) in integrated protection programmes for glasshouse cucumbers. Sbornik UVTIZ, Zahradnictvi 11,215-228 . Landa, Z.,Osborne , L.,Lopez , F.& Eyal,J . (1994).A bioassa y for determining pathogenicity of entomogenous fungi on whiteflies. Biological Control 4, 341-350. Latge, J. P. & Paris, S. (1991). The fungal spore: resevoir of allergens. In The Fungal Spore and Disease Initiation inPlants andAnimals (G.T . Cole &H . C.Hoch ,Eds) , pp. 379-401. Plenum Press, New York. Legaspi, J. C, Nordlund, D. A. & Legaspi Jr., B. C. (1996). Tri-trophic interactions and predation rates in Chrysoperla spp.attackin g silverleaf whitefly. Southwestern Entomologist 21, 33-42. Lenteren, J. C.va n (1995).Integrate d pest management in protected crops. In Integrated pest management (D. Dent, Ed),pp . 311-343. Chapman & Hall, London, UK. Lenteren, J. C. van (2000). A greenhouse without pesticides:Fac t or fantasy? Crop Protection 19, 375-384. Lenteren, J. C. van & Martin, N. A. (1999). Biological control of whiteflies. In Integrated Pest and Disease Management in Greenhouse Crops (R. Albajes, M. L. Gullino, J. C. van Lenteren & Y. Elad, Eds),pp . 202-216. Kluwer Academic Publishers, Dordrecht. Lenteren, J. C. van & Noldus, L. J. J. P. (1990). Whitefly-plant relationships: behavioural and ecological aspects. In Whiteflies: their Bionomics, Pest Status and Management (D. Gerling, Ed), pp. 229-262. Intercept Ltd., Andover, UK. Lenteren, J. C. van & Woets, J. (1988). Biological and integrated pest control in greenhouses.Annual Review of Entomology 33,239-269 . Lenteren, J. C. van, Drost, Y. C, Roermund, H. J. W. van & Posthuma-Doodeman, C. J. A. M. (1997). Aphelinid parasitoids as sustainable control agents in greenhouses. Journal ofApplied Entomology 121, 473-485. Lenteren, J. C. van,Hua ,L iZao ,Kamerman ,J . W. &Rumei ,X u (1995).Th eparasite-hos t relationship between Encarsiaformosa (Hym.,Aphelinidae ) and Trialeurodesvaporariorum (Horn., Aleyrodidae) XXVI. Leaf hairs reduce the capcity of Encarsia to control greenhouse whitefly on cucumber. Journal of Applied Entomology 119, 553-559. Lenteren, J. C. van, Roermund, H.J . W. van & S. Siitterlin, S. (1996).Biologica l control of greenhouse whitefly (Trialeurodes vaporariorum) with theparasitoi d Encarsiaformosa: How does itwork ?Biological control 6, 1-10. Lim,T . K., Ibrahim,Y .B. ,Tang , M. K. & Liew, R. (1991).Occurenc e ofAschersonia placenta and Hypocrella raciborskiio nAsterolecanium ungulatai nDuria n (Diurio zibethinus).Biocontrol Science and Technology 1, 137-144. Lopez-Avila, A. (1986). Chapter 1:Taxonom y and biology. Bemisia tabaci -a literature survey. Louis, I.& Cooke, R. C. (1985).Conidia l matrix and spore germination in some plant pathogens. Transactions of the British Mycological Society 84, 661-667. Lourencao, A.L. ,Yuki ,V . A. &Alves , S.B . (1999).Epizootic s ofAschersonia cf. goldiana on Bemisia tabaci (Homoptera: Aleyrodidae) biotype Bi n the State of SaoPaulo ,Brazil .Anais da Sociedade Entomologica doBrasil 28,343-345 . Macko, V., Staples,R . C, Yaniv, Z. & Granados, R. R (1976). Self-inhibitors of fungal spore germination. In TheFungal Spore: Form andFunction (D.J .Webe r & W. M.Hess ,Eds) ,pp .73-98 .Joh n Wiley & Sons, New York. Mains, E. B. (1959a). North American species of Aschersonia parasitic on Aleyrodidae. Journal of Insect Pathology 1,43-47 . Mains, E. B. (1959b). Species ofAschersonia (Sphaeropsidales). Lloydia 22, 251-221. Mains, E. B. (1959c). species of Hypocrella. Mycopathologia etMycologia Applicata 11, 311-326. Malsam, O., Kilian,M . & Dehne,H .-W .(1998) .Metarhizium anisopliae -Biologica l control of the greenhouse whitefly (Trialeurodes vaporariorum). Bulletin IOBC wprs 2114, 125-128. 154 References Maniania, N.K . &Fargues ,J . (1986).Infectivit y ofPaecilomyces fumosoroseus and Nomuraea rileyi to larvae ofMamestra brassicaean dSpodoptera littoralis a ttw otemperatures .I n Fundamentaland applied aspects ofinvertebrate pathology, Proceedings of theFourth International Colloquiumon Invertebrate Pathology (R.A. Samson, J.M. Vlak & D. Peters, eds.), pp. 251,Veldhoven , The Netherlands. Manzano, M.R. (2000). Evaluation of Amitus fuscipennis as biological control agent of Trialeurodes vaporariorumo n bean in Colombia, Wageningen University, Wageningen, The Netherlands. Marcandier, S. & Khachatourians, G. G. (1987). Susceptibility of the migratory grasshopper, Melanoplus sanguinipes (Fab.) (Orthoptera: Acricicae), to Beauveria bassiana (Bals.) Vuilmin (Hyphomycete): influence of relative humidity. Canadian Entomologist 119, 901-907. Martin,J . H.,Mifsud , D. &Rapisarda , C. (2000). The whiteflies (Hemiptera :Aleyrodidae ) of Europe and the Mediterranean Basin. Bulletin of Entomological Research 90,407-448 . Masuda,T . & Kikuchi,O . (1992).Pathogenicit y of Verticilliumlecanii isolate st owhitefl y andaphids .Japanese Journal ofApplied Entomology and Zoology 36,239-245 . McCoy, C.W . (1985).Citrus :curren t statuso f biological controli nFlorida . InBiological control in agriculture IPM systems (M.A. Hoy & D.C. Herzog, Eds), pp. 481-499. Academic Press, London, UK. McCoy, C.W . (1981).Pes tcontro l byth efungu s Hirsutella thompsonii.I nMicrobial control ofpests and plant diseases, 1970-1980 (H.D.Burges , Ed),pp . 499-512. Academic Press,London , UK. McCoy, C. W. (1990). Entomogenous fungi as microbial pesticides. New Directions in Biological Control, alternativesfor suppressing agricultural pests and diseases. UCLAsymposia on molecular and cellular biology, new series 112, 139-159. McCoy, C. W., Bullock, R. C. & Dybas, R. A. (1982). Avermectin Bl: a novel miticide active against citrus mites in Florida. Proceedings of the Florida State Horticultural Society (publ. 1983) 95, 51-56. McCoy, C. W., Samson, R. A. & Boucias, D. G. (1988). Entomogenous fungi. In CRC Handbook of natural pesticides (CM. Ignoffo & N. Bhushan Mandava, Eds), Vol. vol 5: Microbial insecticides, A: Entomogenous protozoa and fungi, pp.p.151-236 , Boca Raton, Florida. Meade, D. L. & Byrne, D. N. (1991). The use of Verticillium lecanii against submaginal instars of Bemisia tabaci. Journal of Invertebrate Pathology 57,296-298 . Meekes,E . T. M., Fransen, J. J. & Lenteren, J.C. van (1996). Pathogenicity of entomopathogenic fungi of the genusAschersonia against whitefly. IOBC wprs Bulletin 19/1, 103-106. Meekes,E .T .M. ,Voorst , S. van,Joosten ,N .N. ,Fransen , J.J . & Lenteren, J. C. van (2000).Persistenc e of the fungal whitefly pathogenAschersonia aleyrodis, on three different plant species.Mycological Research 104,1234-1240. Merendonk, S. van de & Lenteren, J. C. van (1978). Determination of mortality of greenhouse whitefly Trialeurodes vaporariorum (Westwood) (Homoptera: Aleyrodidae) eggs,larva e andpupa e on four host- plant species:eggplan t (Solanummelongena L.), cucumber (Cucumis sativus L.), tomato (Lycopersicum esculentum L.) and paprika (Capsicum annuum L.). Mededelingen van de Faculteit Landbouwwetenschappen Rijksuniversiteit Gent 43, 421-429. Meyerdirk, D. E., Kreasy, J. B. & Hart, W. G. (1980). Whiteflies (Aleyrodidae) attacking citrus in southern Texas with notes on natural enemies. Canadian Entomologist 112, 1253-1258. Milner, R. J. (1997). Prospects for biopesticides for aphid control.Entomophaga 42, 1-2, 227-239. Milner, R. J. & Lutton, G. G. (1986). Dependence of Verticillium lecanii (Fungi: Hyphomycetes) on high humidities for infection and sporulation using Myzus persicae (Homoptera: Aphididae) as host. Environmental Entomology 15,380-382 . Minks, A. K. & Harrewijn, P. (1987). Aphids. Their biology, Natural Enemies and Control. Elsevier, Amsterdam. Moore,D. , Higgins,P .M . &Lomer ,C .J . (1996). Effects of simulted and natural sunlight on the germination of conidia ofMetarhiziumflavoviride Gams andRozysypa l andinteraction s with temperature. Biocontrol Science and Technology 6, 63-76. Moore, D. & Prior, C. (1993). The potential of mycoinsecticides. Biocontrol News and Information 14, 31N- 40N. 155 References Mor, H., Gindin, G., Ben-Ze'ev, I. S., Raccah, B., Geschtovt, N. U. , Ajtkozhina, N. & Barash, I. (1996). Diversity among isolates of Verticillium lecanii as expressed by DNA polymorphism and virulence towards Bemisia tabaci. Phytoparasitica 24, 111-118. Mora Morin,J . (1985).Distributio n of the natural enemies of citrus white flies in Cuba (part 1).Agrotecnia de Cuba 17, 15-25. Morrill, A. W. & Back, E. A. (1912). Natural control of whiteflies in Florida. United States Department of Agriculture, Bureau of Entomology Bulletin 102, 1-78. Motte, G. & Unger, J. G. (1995).Zu mAuftrete n von EchtemMehlta u (Oidiumspp. ) an Poinsettien (Euphorbia pulcherrimi) in Danmark.Nachrichtenblatt des Deutschen Pflanzenschutzdienstes 47,22 . Muraleedharan, N.(1985) .A new disease of the cardamom whitefly. Tropical Pest Management 31, 234-235. Neal,J . W., Leonhardt, B.A. ,Brown , J. K.,Bentz ,J . & Devilbiss,E .D . (1994).Cuticula r Lipids of greenhouse whitefly and Sweetpotato whitefly type A and B (Homoptera: Aleyrodidae) pupal exuviae on the same hosts.Annals of the Entomological Society ofAmerica 87,609-618 . Nene,Y .L . (1973).Not eo n a fungus parasiteo f Bemisia tabaci Genn., a vector of several plant viruses. Indian Journal ofAgricultural Sciences 43,514-516 . Nguyen, R. & Hamon, A. (1985). Cloudy-winged whitefly, Dialeurodes citrifolii (Morgan) (Homoptera: Aleyrodidae: Aleyrodinae). Entomology Circular, Division of Plant Industry, Florida Department of Agriculture and Consumer Services, 2 pp. Obornik,M. , Stouthamer,R. , Meekes,E . & Schilthuittzen,M . (1999).Molecula r characterization and phytogeny of the entomopathogenic fungus Aschersonia spp.Plant Protection Science 35, 1-9. Oetting, R. D. & Buntin, G. D. (1996). Bemisia damage expression in commercial greenhouse production. In Bemisia 1995:Taxonomy, Biology, Damage Control and Management (D.Gerlin g & R.T. Mayer, Eds), pp. 201-208.Intercep t Ltd., Andover, UK. Oho, N. (1967). Possible utilization of Aschersonia aleyrodis for control of citrus whitefly Dialeurodes citri Ashmead. Proceedings of thejoint United states Japan seminar on microbial control of insect pests, Fukuoka, april, 119-128. Onillon, J. C. (1990). The use of natural enemies for the biological control of whiteflies. In Whiteflies: their bionomics, pest status and management (D. Gerling, Ed),pp . 147-186.Intercep t Ltd., Andover, UK. Osborne, L. S. & Landa, Z. (1992). Biological control of whiteflies with entomopathogenic fungi. Florida Entomologist 74,456-471. Osborne,L .S. ,Hoemler ,K . & Gerling,D .(1990a) .Prospect sfo r biological control of Bemisia tabaci. Bulletin WBCwprs 13/5, 153-160. Osborne,L . S., Storey, G. K.,McCoy , C. W. & Walter, J. W. (1990b).Potentia l for controlling the sweetpotato whitefly, Bemisia tabaci, with the fungus, Paecilomyces fumosoroseus. In Proceedings and Abstracts, Fifth International Colloquium on Invertebrate Pathology and Microbial Control, pp. 386, Adelaide, Austalia. Parbery, I. H., Brown, J. F. & Bofinger, V. J. (1981). Statistical methods in the analysis of phylloplane populations. In Microbial ecology of thephylloplane (J. P. Blakeman, Ed), pp. 47-65. Academic Press, London, UK. Parsana,G . J., Butani, P.G . &Kapadia ,M .N . (1995). Preference of the sugarcane whitefly and its pathogenic fungus to different promising varieties.Indian Sugar, 239-241. Pas, R. K. van der, Ravensberg, W. J., Braver, A. C, Buijsen, A. C. van & Malais, M. (1996). Comparison between MYCOTAL (Verticilliumlecanii) an dAschersonia aleryodisfo r thecontro l of whitefly. Bulletin IOBC wprs 19/9,200-203 . Pas, R. K. van der, Ravensberg, W. J. & Cryer, E. (1998). Insect pathogenic fungi for environmental-friendly pest control in the glasshouse. 'Investigating oil formulations'. Bulletin IOBC wprs 21/4, 129-132. Paulson, G. S. & Beardsley, J. W. (1985). Whitefly (Hemiptera: Aleyrodidae) egg pedicel insertion into host plant stomata.Annals of the Entomological Society ofAmerica 78,506-508 . Payne,R .W. ,Lane ,P .W. , Ainsley,A .E. ,Gower ,J . C, Tunnichff Wilson,G . & Paterson,L .J . (1987).Genstat 5: Reference Manual. Clarendon Press, Oxford, UK. Pekrul, S. & Grula,E .A . (1979).Mod e of infection of thecor n earworm (Heliothis tea) byBeauveria bassiana asreveale d by scanning electron microscopy.Journal ofInvertebrate Pathology 34,238-247 . 156 References Petch, T. (1921). Sudies in entomogenous fungi: II.- The genera Hypocrella and Aschersonia. Annals of the Royal Botanical Garden, Peradeniya 7, 169-279. Petch, T. (1925). Entomogenous fungi: additions and corrections. Transactions of the British Mycological Society 10, 190-210. Petch,T . (1932).Note s on entomogenous fungi. Transactions of the British Mycological Society 16,209-245 . Petch, T. (1939).Note s on entomogenous fungi. Transactions of the British Mycological Society 23, 127-148. Pettersson, M. L. (1993). Horticultural pests and diseases in 1993. Vaxtskyddsnotiser 58,21-26 . Ponomarenko, N. G., Prilepskaia, N. A., Murvanidze, M. J. A. & Stolyarova, L. A. (1975).Aschersonia for control of thecitru s whitefly Dialeurodes citri. Zashchita Rastenii, (no.6 )44-45 . Poprawski, T. J., Greenberg, S. M. & Ciomperlik, M. A. (2000). Effect of host plant on Beauveria bassiana- and Paecilomyces fumosoroseus-induced mortality of Trialeurodes vaporariorum (Homoptera : Aleyrodidae). Environmental Entomology 29, 1048-1053. Price,P .W. ,Bouton , C. E., Gross, P.,McPheron , B. A., Thompson, J. N. & Weiss, A. E. (1980). Interactions among three throphic levels: Influence of plants on interactions between insect herbivores and natural enemies.Annual Review of Ecology and Systematics 11,41-65 . Primak, T. A. & Chizhik, R. I. (1977). The possibilities of using the fungus Aschersonia for control of the greenhouse whitefly. Zakhist Roslin 22, 53-56. Prior, C. (1992). Discovery and characterization of fungal pathogens for locust and grasshopper control. In Biological control of Locusts and Grasshoppers (C.J. Lomer & C. Prior, Eds), pp. 159-180. CAB International, Wallingford, UK. Protsenko, E.P .(1959) .Th e importance of the fungus Aschersonia in nature and its practical use by man in the biological control of insects.Sb. Karantinu Rast 19, 147-213. Quezada, J.R . (1974).Biologica l control ofAleurocanthus woglumi [Homoptera: Aleyrodidae] in El Salvador. Entomophaga 19,243 -254 . Ramakers, P. M. J. (1983).Aschersonia aleyrodis, a selective biological insecticide. Bulletin IOBC wprs 6/3, 167-171. Ramakers, P. M. J., Rombach, M. C. & Samson, R. A. (1982). Application of the entomopathogenic fungus Aschersonia aleyrodis in an integrated control programme against the glasshouse whitefly Trialeurodes vaporariorum. In Third International Colloquium on Invertebrate Pathology and Microbial Control, University of Sussex, Brighton, UK. Ramakers, P. M. J. & Samson, R. A. (1984). Aschersonia aleyrodis, a fungal pathogen of whitefly. II Application as abiologica l insecticide in glasshouses.Journal ofApplied Entomology 97, 1-8. Ramoska, W. A. (1984). The influence of relative humidity onBeauveria bassiana infectivity and replication in the chinch bug, Blissus leucopterus. Journal ofInvertebrate Pathology 43, 389-394. Ramoska,W . A. &Todd ,T . (1985).Variatio n inefficac y andviabilit y ofBeauveria bassiana in thechinc h bug (Hemiptera: Lygaeidae) as aresul t of feeding on selected host plants.Environmental Entomology 14, 146- 148. Rao, N. V., Reddy, A. S. & Rao, K. T. (1989).Natura l enemies of cotton whitefly, Bemisia tabaci Gennadius inrelatio n to host population and weather factors. Journal of Biological Control 3, 10-12. Ravensberg,W . J., Malais,M . &Schaaf ,D . A. van der (1990). Application of Verticillium lecanii in tomatoes and cucumber to control whitefly and thrips.Bulletin IOBC wprs 13/5,173-178. Riba, G. & Entcheva, L. (1984). Influence de l'hygrometrie ambiante sur l'agressivite comparee de plusieurs Hyphomycetes entomopathogenesal'egar d del'Aleurod e desserres ,Trialeurodes vaporariorum (Westw.) . Comptes Rendus des Seances de lAcademie dAgriculture de France 70, 521-526. Risch, S. J. (1985). Effects of induced chemical changes on interpretation of feeding preference tests. Entomologia Experimental etApplicata 39, 81-84. Roberts,D .W . (1996).Th erol e ofbaci sresearc h onth edevelopmen t of fungi asbioinsecticides . In VSiconbiol: Simposio de Controle Biologico, pp. 108-118, Parana, Brasil. Roberts, D. W. & Humber, R. A. (1981). Entomogenous fungi. In Biology of conidialfungi (G.T. Cole &B . Kendrick, Eds), Vol. 2, pp. 201-236. Academic Press, Berlin. 157 References Robertson, J. L., Preisler, H. K., Ng, S. S., Hickle, L. A. & Gelernter, W. D. (1995). Natural variation: a complicating factor in bioassays with chemical and microbial persticides. Journal of Economic Entomology 88, 1-10. Roennund, H. J. W. (1995). Understanding biological control of greenhouse whitefly with the parasitoid Encarsia formosa. From induvidual behaviour to population dynamics, Wageningen Agricultural University, Wageningen. Roermund, H. J. W. & Lenteren, J. C. van (1992). Life-history parameters of the greenhouse whitefly, Trialeurodes vaporariorum and the parasitoid Encarsia formosa. Wageningen Agricultural University Papers 92-3, 147pp . Rombach, M. C. & Gillespie, A. T. (1988). Entomogenous hyphomycetes for insect and mite control on greenhouse crops.Biocontrol News and Information 9, 7-18. Rombach, M.C . & Samson, R. A. (1982).Smal l scaleproductio n of theentomopathogeni c fungus Aschersonia aleyrodis.I nThird International Colloquiumon Invertebrate Pathology and Microbial Control,Universit y of Sussex, Brighton, UK. Rovesti, L., Grazzi, G. & Viccinelli, R. (1997). Use of entomopathogenic fungi for pest control in protected crops in Italy.Bulletin IOBC wprs 20/4, 285-289. Sabelis, M. W., Baalen, M. van, Bakker, F. M., Bruin, J., Drukker, B.,Egas , M., Janssen, A. R. M., Lesna, I. K., Pels,B. ,van ,Rij n P.C .J . & Scutareanu, P.(1999) .Th e evolution of direct and indirect plant defence against herbivorous arthropods. In Herbivores: Between Plants and Predators (H. Olff, V. K. Brown & R. H. Drent, Eds),pp . 109-166.Blackwel l Science,Oxford , UK. Sabelis,M .W . &Jong , M.C . M.d e (1988).Shoul d allplant s recruit bodyguards?Condition s for a polymorphic ESS of synomone production in plants. Oikos 53,247-252 . Samson,R . A. (1995).Constraint s associated with taxonomy of biocontrol fungi. Canadian Journal of Botany 73 (Suppl. 1), S83-S88. Samson, R. A. & McCoy, C. W. (1983). Aschersonia aleryodis, a fungal pathogen of whitefly. I. Scanning electron microscopy of the development on the citrus whitefly. Journal ofApplied Entomology 96,380 - 386. Samson, R. A. &Rombach , M.C . (1985).Biolog y of thefung i Verticillium andAschersonia. In Biological pest control, theglasshouse experience(N.W .Husse y & N. Scopes,Eds) ,pp . 34-42.Blandfor d Press, Dorset, UK. Samson, R. A., Rombach, M. C, Evans, H. C. & Riba, G. (1982). Comparative studies in vitro on various specieso f theentomopathogeni c genusAschersonia. In Third International Colloquium on Invertebrate Pathology andMicrobial Control, University of Sussex, Brighton, UK. Sanada, W. T., Pulz, F. S., Murai, N. T., Idagawa, T., Bleicher, E. & Mariconi, F. A. M. (1974). A new experiment for the control of the "woolly whitefly" Aleurothrixus floccosus. Anais de Sociedade Entomologica do Brasil 3, 14-19. Santamaria, A., Costa-Comelles, J., Alonso, A., Rodriguez, J. M. & Ferrer, J. (1998). Trials with the entomopathogenic fungus Beauveria bassiana (Balsamo) Vuillemin for thecontro l of theWooll y Whitefly Aleurothrixus floccosus (Maskell) (Homoptera: Aleyrodidae) and effect on the parasitoid Cales noacki (Howard) (Hymenoptera: Aphelinidae). Boletin de Sanidad Vegetal,Plagas 24, 695-706. Schaaf,D .A . van der,Malais ,M .& Ravensberg, W.J . (1990). The use of Verticillium lecanii against whitefly and thrips inglasshous e vegetables in theNetherlands . In Proceedings andAbstracts, Fifth International Colloquium on Invertebrate Pathology and Microbial Control, pp. 391, Adelaide, Australia. Schaaf,D .A . van der, Ravensberg, W. J. & Malais,M . (1991). Verticillium lecanii as a microbial insecticide against glasshouse whitefly. Bulletin IOBC wprs 14/7, 120-123. Schneider, R. (1954). Untersuchungen iiber Feuchtigkeitsanspriicheparasitische r Pilze. Phytopathologischer Zeitschrift 21, 63-78. Schreiter, G., Butt, T. M., Beckett, A., Vestergaard, S. & Moritz, G. (1994). Invasion and development of Verticilium lecaniii n thewester n flower thripsFrankliniella occidentalis. Mycological Research 98, 1025- 1034. 158 References Schultz, J. C. & Keating, S. T. (1991). Host-plant-mediated interactions between the gypsy moth and a baculovirus.I nMicrobial Mediation of Plant-Herbivore Interactions (P.Barbosa , V.A. Krischik & C.G. Jones, Eds),pp . 489-506. John Wiley & Sons,Ne w York, USA. Schuster, D.J. , Stansly,P .A . &Polston , J.E .(1996) .Expressio n of plant damageb yBemisia .I nBemisia 1995: Taxonomy, Biology, Damage Control and Management (D. Gerling & R.T. Mayer, Eds),pp . 153-165. Intercept Ltd., Andover, UK. Sebastian, S.& Gupta,P .(1996) .Evaluatio n of three plant latices as seed treatment againstPratylenchus thornei on chickpea. International Chickpea and Pigeonpea Newsletter 3, 39-40. Selvakumaran, S., David, B. V. & Kumaresan, D (1996).Observation s on the natural enemies of the whitefly Kanakarajiellacardamomi (Davi dan dSubramaniam )a pes to ncardamom .Indian Journal ofEnvironment and Toxicology 6, 26-27. Shelton, A. M. & Roush, R. T. (2000). Resistance to insect pathogens and strategies to manage resistance. In Fieldmanual of techniques in invertebrate pathology: applicationand evaluation of pathogens for control of insectsand other invertebratepests (L.A . Lacey &H . K. Kaya, Eds),pp . 829-845.Kluwe r Academic Publishers, Dordrecht, The Netherlands. Shelton, A.M. ,Vandenberg , J. D.,Ramos , M. & Wilsey, W. T. (1998). Efficacy and persistence of Beauveria bassianaan dothe rfung i for controlo f diamondbackmot h (Lepidoptera: Plutellidae)o ncabbag eseedlings . Journal of Entomological Science 33, 142-151. Shimizu, S. & Aizawa, K. (1988). Nature of blastospores produced by an enthomopathogenic fungus, Paecilomyces fumosoroseus. Japanese Journal ofApplied Entomology and Zoology 32, 182-186. Shu, ZhengYi (1996). A study of the biology of Aschersonia aleyrodis Webber. Journal of Southwest Agricultural University 18,524-526 . Sinadskii, Yu V. & Kozarzhevskaya, E. F. (1980). Biological protection of flowering and decorative plants. Zashchita Rastenii, (no. 10) 21-22. Sinha, S. (1971). The microflora on leaves of Capsicum annuum (L.) Watt E.D., Solanum melongena L., Solanum tuberaosumL . andLycopersicum esculentum Mill. In Ecology of leaf surface micro-organisms (T.F. Preece & C.H. Dickinson, Eds),pp . 175-189.Academi c Press, London, UK. Smith,P . (1994). Increased infectivity of oil and emulsifiable oil formulations of Paecilomyces fumosoroseus conidia toBemisia tabaci. In Sixth International Colloquium on Invertebrate Pathology and Microbial Control,pp . 82,Montpellier , France. Smith, I. M., Dunez, J., Phillips, D. H. ,Lelliott , R. A. & Archer, S. A. (1988). European handbook of plant diseases. Blackwell Scientific Publications, Oxford, UK. Smith-Kielland,I. ,Dornish , J.M. ,Malterud , K. E.,Hvistendahl , G.,R0mming , Chr.,B0ckman , O.C , Kolsaker, P., Stenstr0m,Y . &Nordal , A. (1996).Cytotoxi c triterpenoids from theleave s of Euphorbia pulcherrima. Planta Medica 64,322-325 . Solovei, E. F. (1981).Aschersonia in greenhouses. Zashchita Rastenii, (no. 1) 30-31. Soper, R. S. & Ward, M. G. (1981). Production, formulation and application of fungi for insect control. In Biological controlin crop protection, (G.C.Papavizas , Ed),pp . 161-180. Allenheld & Osmun Publ. Co., Montclair, NY, USA. Spassova, P. (1980). Investigation on the entomopathogenic fungus Aschersonia sp. Gradinarska i Lozarska Nauka 17, 77-83. Spassova, P., Khristova, E. & Elenkov, E. S. (1980). Pathogenicty of various species of fungi on the genus Aschersonia tolarva e of the greenhouse whitefly (Trialeurodes vaporariorum Westw.) on tomatoes and cucumbers. Gradinarska iLozarska Nauka 17,70-75 . St.-Leger, R. J. (1991). Integument as a barrier to microbial infections. In Physiology of the Insect Epidermis (K. Binnington & A. Retnakaran, Eds),pp . 284-306. CSIRO Publications, Melbourne, Australia. St.-Leger, R. J., Goettel,M. ,Roberts , D.W . & Staples,R . C. (1991).Penetratio n events during infection of host cuticle byMetarhizium anisopliae. Journal of Invertebrate Pathology 58, 168-179. Stamp, N. E. & Bowers, M. D. (1994). Effects of cages, plant age and mechanical clipping on plantain chemistry. Oecologia 99,66-71. Steenberg, T. & Humber, R. A. (1999). Entomopathogenic potential of Verticillium and Acremonium species (Deuteromycotina: Hyphomycetes). Journal of Invertebrate Pathology 73,309-314 . 159 References Steenis, M. van (1995). Evaluation and application of parasitoids for biological control of Aphis gossypii in glasshouse cucumber crops, Wageningen Agricultural University, Wageningen. Steinhaus, E. A. (1958). Stress as a factor in insect disease. In Proceedings 10th International Congress of Entomology, Vol. 4, pp. 725-730. Mortimer Ltd., Ottawa, Canada, Montreal. Steinkraus, D. C, Oliver, J. B., Humber, R. A. & Gaylor, M. J. (1998). Mycosis of bandedwinged whitefly (Trialeurodes abutilonea) (Homoptera: Aleyrodidae) caused by Orthomyces aleyrodis gen. & sp. nov. (Entomophfhorales: Entomophthoraceae).Journal of Invertebrate Pathology 72, 1-8. Steyn, P. S. (1995). Mycotoxins, general view, chemistry and structure. Toxicology Letters 82-83, 843-851. Strasser, H., Abendstein, D., Stuppner, H., Schinner, F. & Smits, P. H. (1998). Detection of primary and secondary metabolites excreted byBeauveria brongniartii. Bulletin IOBC wprs21/4 , 61-66. Strasser, H., Abendstein, D., Stuppner, H. & Butt, T. M. (2000a). Monitoring the distribution of secondary metabolites produced by the entomogenous fungus Beauveria brongniartii with particular reference to oosporein. MycologicalResearch 104, 1227-1233. Strasser, H., Vey, A. & Butt, T. M. (2000b). Are there any risks in using entomopathogenic fungi for pest control, with particular reference to the bioactive metabolites of Metarhizium, Tolypocladium and Beauveria species? Biocontrol Science and Technology 10,717-735 . Summers, C. G., Newton Jr., A. S. & Estrada, D. (1996). Intraplant and interplant movement of Bemisia argentifolii (Homoptera: Aleyrodidae) crawlers.Environmental Entomology 25, 1360-1364. Sussman,A . S. (1968).Longevit y and survivability offungi . In Thefungi: anadvanced treatise,volume III: The fungal population (G.C. Ainsworth & A.S. Sussman, Eds), Vol. 3:Th e fungal population, pp. 447-486. Academic Press,London , UK. Siitterlin, S. (2000). Biological Control of Whitefly on Gerbera: Success or Failure, Wageningen University, Wageningen. Siitterlin, S. & Lenteren,J . C. van (1996). Hairiness of Gerberajamesonii leaves and the walking speed of the parasitoid Encarsiaformosa. Bulletin IOBC wprs 19/1, 171-174. Siitterlin, S. &Lenteren , J. C. van (1997). Influence of hairiness of Gerberajamesonii leaves on the searching efficiency of the parasitoid Encarsiaformosa. Biological Control 9, 157-165. Sutton, B.C . (1980). The Coelomycetes, Fungi Imperfecti withpycnidia, acervuli and stromata. Szabo, P., Lenteren, J. C. van & Huisman, P. W. T. (1993). Development time, survival and fecundity of Encarsiaformosa onBemisia tabaci and Trialeurodes vaporariorum.Bulletin IOBC wprs 16/2, 173-176. Tallamy, D. W., Whittington, D. P., Defurio, F., Fontaine, D. A., Gorski, P. M. & Gofhro, P. W. (1998). Sequestered cucurbitacins and pathogenicity of Metarhizium anisopliae (Moniliales: Moniliaceae) on spotted cucumber beetle eggs and larvae (Coleoptera: Chrysomelidae). Environmental Entomology 27, 366-372. Tanada, Y.& Kaya,H . K. (1993).Hos tresistance .I nInsect Pathology (Y.Tanad a & H.K. Kaya, Eds),pp .492 - 553. Academic Press Inc.,London , UK. Thomas,D .C . (1993).Hos tplan tadaptatio n inth eglasshous e whitefly. VI.Ca n aDutc h straino f whitefly adapt to a Dutch strain of sweet pepper?Journal ofApplied Entomology 115,405-415 . Thumar, R. K. & Kapadia, M. N. (1994). Ovicidal control of Aleurolobus barodensis (Maskell) and its suppression by the entomophagous fungus. Indian Sugar 44, 501-502. Uchida, M. (1970). Studies on the use of the parasitic fungus Aschersonia sp.fo r controlling citrus whitefly, Dialeurodes citri. Bulletin of the Kanagawa Horticultural Experiment Station, 66-74. Uziel, A. & Kenneth, R. G. (1991). Survival of primary conidia and capilloconidia at different humidities in Erynia (subgen. Zoophfhora) spp. and inNeozygites fresenii (Zygomycotina: Entomophthorales), with special emphasis on Erynia radicans. Journal of Invertebrate Pathology 58, 118-126. Vandenberg, J. D., Shelton, A. M., Wilsey, W. T. & Ramos, M. (1998). Assessment of Beauveria bassiana sprays for control of diamondback moth (Lepidoptera: Plutellidae) on crucifers. Journal of Economic Entomology 91, 624-630. Vargas Sarmiento,M .M. , Rodriguez Sierra, D.A. , Sanabria, J. &Lopez-Avila ,A . (1995).Ensay o de diferentes dosis de Aschersonia aleyrodis Webber y parasitismo de Encariaformosa Gahan en ninfas de tercer y cuarto instar de la mosca blanca de los invernaderos. Revista Colombiana de Entomologia 21, 159-170. 160 References Verhaar, M. A. (1998). Studies in biological control of powdery mildew in cucumber (Sphaerotheca fulginea) andros e (S.pannosa ) bymean s of mycoparasites,Wageninge n Agricultural University, Wageningen, The Netherlands. Verhaar, M. A., Hijwegen, T. & Zadoks, J. C. (1998). Selection of Verticillium lecanii isolates with high potential for biocontrol of cucumber powdery mildew by means of components analysis at different humidity regimes.Biocontrol Science and Technology 8, 465-477 . Vet, L. E. M. & Dicke, M. (1992). Ecology of infochemical use by natural enemies in a tritrophic context. Annual Review of Entomology 37, 141-172. Vidal, C, Lacey, L.A . & Fargues, J. (1997). Pathogenicity of Paecilomyces fumosoroseus (Deuteromycotina: Hyphomycetes) against Bemisia argentifolii (Homoptera: Aleyrodidae) with a description of a bioassay method. Journal of Economic Entomology 90,765-772 . Vidal, C, Osborne, L. S., Lacey, L. A. & Fargues, J. (1998). Effect of host plant on the potential of Paecilomyces fumosoroseus (Deuteromycotina: Hyphomycetes) for controlling the silverleaf whitefly, Bemisia argentifolii (Homoptera: Aleyrodidae) in greenhouses. Biological Control 12, 191-199. Vogelezang, J. V. M. (1993). Bench heating for potplant cultivation: analysis of effects of root - and air temperature on growth, development anproduction , Wageningen Agricultural University, Wageningen, The Netherlands. Walstad, J.D. ,Anderson , R. F. & Stambaugh,W . J. (1970).Effect s of environmentalcondition s on two species of muscardinefung i (Beauveria bassianaan dMetarhizium anisopliae).Journal ofInvertebrate Pathology 16, 221-226. Watanabe, M. A., Yoshii, C. & Siloto, R. C. (1994). Parasitism of the scale insect Selenaspidus articulatus(Morgan, 1889) (Hemiptera/Homoptera, Diaspididae) on Citrus in the regions of Jaguariuna and Limeira, SP.Revista deAgricultura Piracicaba 69, 193-200. Watson,J . R. (1912).Th e "naturalmortality " of the whitefly. Florida Agricultural Experiment Station, Annual Report 1912, xlviii-lxiii. Weber, H. (1931). Lebensweipe und Umweltbeziehunen von Trialeurodes vaporariorum (Westwood) (Homoptera: Aleyrodidae). Zeitschriftfur Morphologie und Okologie der Tiere 23, 575-753. Wolcott, G. N. (1955).Entomogenou s fungi in Puerto Rico. Science 121,875-876 . Wraight, S. P. & Bradley, C. A. (1996). Production, formulation, and application technologies for the use of entomopathogenic fungi to control field crop pests. In VSiconbiol: Simpdsio de Controle Biologico, pp. 170-177,Parana , Brasil. Wraight,S .P .& Carruthers, R. I. (1999).Production , delivery, andus eo f mycoinsecticidesfo r control of insect pestso nfield crops .I nBiopesticides: Useand Delivery (J. J. Menn & F.R . Hall,Eds) ,Vol . 5,pp .23 3 - 269. Humana Press,Totowa , NJ. Wraight,S .P. , Carruthers,R . I., Bradley,C .A. , Jaronski, S. T., Lacey, L. A., Wood, P. & Galaini-Wraight, S. (1998).Pathogenicit y of theentomopathogeni c fungi Paecilomyces andBeauveria bassiana. against the silverleaf whitefly, Bemisia argentifolii. Journal ofInvertebrate Pathology 71, 217-226. Wraight, S. P. , Butt, T. M. , Galaini-Wraight, S. , Allee, L. L. , Soper, R. S. & Roberts, D. W. (1990). Germination and infection processes of the entomophthoralean fungus Erynia radicans on the potato leafhopper, Empoascafabae. Journal of Invertebrate Pathology 56, 157-174. Wraight, S. P., Carruthers, R. I., Jaronski, S. T., Bradley, C. A., Garza, C. J. & Galaini-Wraight, S. (2000). Evaluation of the entomopathogenic fungi Beauveria bassiana and Paecilomyces fumosoroseus for microbial control of the silverleaf whitefly, Bemisia argentifolii. Biological Control 17,203-217 . Yasui, M., Fukada, M. & Maekawa, S. (1985). Effects of buprofezin on different development stages of the greenhouse whitefly, Trialeurodes vaporariorum (Westwood) (Homoptera: Aleyrodidae). Applied Entomology and Zoology 20, 340-347. Zibul'skaya, G. N., Czizik, P. I. & Scatinskaya, E. I. (1975). Cucumber protection in greenhouses. Zashchita Rastenii, (no. 5) 25-26. Zou,W . H.(1988) .Parasitis mo fDialeurodes citriby Aschersoniaplacenta. ZhejiangAgricultural Science,280 - 282. Zouba, A. & Kahn,A .J . (1992). Verticillium lecanii a natural parasite on the fig rust Ceratoliumfici. Arabian Journal of Plant Protection 10, 34-31. 161 References 162 Summary Many horticultural and agricultural crops are good host plants for the greenhouse whitefly, Trialeurodes vaporariorum, and the silverleaf whitefly, Bemisiaargentifolii. Thei r damage tocrop si smanifold . Whenpresen ti n sufficient numbers they can cause leaf drop and inhibit fruit maturation.The yar eefficien t vectorso feconomicall yimportan tplan tviruses . Whiteflies produce honeydew, which soils and damages crops andhamper s theprocessin g of products sucha scotton .I tserve sa sa substrat efo r sootymoulds ,thu sreducin glea f photosynthesis and rendersplant s andfruit s unsightly. For export of plantso rplan tproduct s azero-toleranc e is valid, especially when theinsec t is aquarantin e pest in the importing country, therefore the simplepresenc e of whitefly can leadt odestructio n orre-expor t of anentir e shipment. As more environmentally responsible agricultural strategies are adopted, natural enemieso fbot h whitefly species,wil lpla y anincreasin g role. Screening for natural enemies which areabl et okil lbot hpes tinsect squickly , without affecting other natural enemies, isa n important line of research. Entomopathogenic fungi can meet these requirements and can therefore be a valuable asset to existing biological and chemical control measures. Our attentioni sdirecte d towardsth egenu sAschersonia, because previous research indicated that Aschersonia aleyrodis isa promisin gwhitefl y controlagen tbecaus eo fit stoleranc et orelativ e humidities aslo w as50% ,it slon gpersistenc e onlea f surfaces and itscompatibilit y with the parasitoid Encarsiaformosa. This project consisted of two components: 1) to identify virulent isolates of Aschersonia spp.fo r theus eagains tgreenhous e andsilverlea f whitefly, and2 )t ostud y factors whichinfluenc eth eeffectivit y ofentomopathogeni cfungi ,wit hspecia lreferenc et ohos tplant , humidity and their interaction. Entomopathogenic fungi of the genus Aschersonia are specialized on whitefly and scale insects and can be used as a biological control agent against silverleaf whitefly, B. argentifolii and greenhouse whitefly, T. vaporariorum. In chapter 2 44 isolates of Aschersonia spp. were tested for their ability to sporulate and germinate on semi-artificial media and to infect the insect host. Seven isolates sporulated poorly (less than 5.107 conidia/culture)an dte nwer eno tabl et oinfec t either of the whitefly species. Several isolates wereabl et oproduc ecapilliconidi ao nwate ragar .Th eInfectio n levelwa sno tcorrelate d with germinationo nwate ragar . After aselectio nbase do nspor eproductio n andinfection , virulence of 31 isolates was evaluated on third instar nymphs of both whitefly species on poinsettia {Euphorbia pulcherrima). Infection levels varied between 2% to 70%, and infection percentageso fB. argentifolii correlate dwit htha to fT. vaporariorum. However,mortalit ywa s 163 Summary higherfo rT. vaporariorum than for B. argentifolii, asa resul t of ahighe r 'mortality byothe r causes' of greenhouse whitefly on poinsettia. Several isolates, among which unidentified specieso fAschersonia originatin gfro m Venezuelaan dMalaysia ,A. aleyrodisfro m Colombia, and A.placenta from India showed consistent results in their ability to infect both whitefly species. Out of the above 44 isolates, six isolates were selected based on virulence, spore production and geographical origin. In chapter 3 the LD50 and LT50 values ofA .aleyrodis, A. placenta and four Aschersonia spp. isolates were compared to select the most virulent isolate for control of B. argentifolii and T. vaporariorum.Bioassay s were carried out on predominantly second instarnymph so nintac tpoinsetti aplant sunde rglasshous e conditions andth edosag erange dfro m 14t o 1.4 x 105conidia/cm 2.Fiv e isolates werehighl y pathogenic tobot hwhitefl y species;mortalit yb yon eisolat e didno texcee d 50%,therefor e noLD 50'so r 2 LT50'swer e calculated. OnB. argentifoliiLD 50'svarie d between 1600an d480 0conidia/cm andLT 50'sbetwee n4. 6- 8. 7day san do nT. vaporariorumLD 50'svarie dbetwee n70 0an d530 0 2 conidia/cm and LT50's between 4.5 and 9.9 days. LT50's based on visual signs are not very accuratefo r sedentaryinsects ,lik ewhitefl y nymphs,therefor e we also looked atth enympha l instari nwhic hth einfecte d nymphsdied .Fo rfou r isolatesth emajorit y ofnymph sdie di n the first to third instar period. However, the differences found between isolates did not always agreewit hdifference s found comparingLT 50's.Fo rtw oisolate sth eoptimu mmortalit ydi dno t occur atth e highest dosage, which was alsoreflecte d in the speed ofkill . In chapter 4 scanning electron microscopy and bioassays were carried out to obtain better insight into the infection process of A. aleyrodis,A. placenta and an unidentified Aschersoniasp. .Conidi ao fAschersonia spp .germinate dreadil yo nth ecuticul ao fhos tinsect s as well as on water agar. On water agar A. placenta also produced capilliconidia. No germination was observed on poinsettia leaf surface, except on the leaf veins. On B. argentifoliith efung i formed largeamount so fmucilag et oattac hthemselve s toth einsect . Appressoria wereforme d before penetration, but also direct penetration was observed. This seemedno trelate dt oa specifi c siteo nth einsect .Fo rbot hwhitefl y species,firs t tothir dinsta r nymphswer emos tsusceptible .I fth epopulatio n existedo ffourt h instarnymph sfo r moretha n 50%,infectio n levels dropped from 90 to50% .Infecte d whitefly nymphsusuall y diedi n the stagefollowin g thetreate d stage.Th efungu s protrudedfro m theinsec t viath emargin s or via theemergenc efold s ofpupae ,i fhumidit ylevel swer ehig henough .However , sporulationwa s rarelyobserve do nwhiteflie s developingo npoinsetti aplant sunde rth eprevailin gconditions . Persistence ofA .aleyrodis was studiedi n chapter 5o ncucumbe r (Cucumis sativus), gerbera{Gerbera jamesonii) an dpoinsettia . Germination capacity and infectivity of conidia, which stayedo nth edifferen t plants overa perio do fu pt oon emonth , wereassessed .Averag e 164 Summary germination of conidia on theleave s was low (< 14.3%),wherea s itwa s shown thatmos to f theconidi atransferre d from thelea ft owate r agarwer eviable ,eve n after having been present onth elea f surface for onemonth .Germinatio n capacitywa sinfluence d byhos tplan t species: it was highest on cucumber, followed by poinsettia and lowest on gerbera. On cucumber leaves,conidi a stayedviabl e andwer e ablet oinfec t 90%o f thewhitefl y nymphs,eve n at 31 days after spore application. On gerbera, germination capacity decreased considerably from 80% (day0 )t o40 % (day 31).Thi s was reflected innympha l mortality, which declined from 75% to 40%.Despit e thehig h germination capacity (60%) of conidia on poinsettia after an exposureo fon emonth ,nympha lmortalit ydecrease dfro m 70%a tth eda yo f sporeapplicatio n to 10%afte r three days atlea f surface, andremaine d lowthroughou t themonitorin g period. Thephyllospher emicroflora , secondaryplan tmetabolite san dmicroclimat eca npla ya rol ei n these findings. Can phyllosphere humidity explain differences in insect mortality due to entomopathogenic fungi on different crops? This was tested in chapter 6 for cucumber, gerbera,tomat oan dpoinsetti a onwhic hmycosi so fT. vaporariorum and cotton aphid {Aphis gossypii) were determined in relation to host-plant climate. Phyllosphere humidity was estimated using climate and host-plant parameters. Hydrophobicity of the leaves and crop densitywer eals otake nint oaccount .O ncucumber , gerbera andtomato ,th efung i A.aleyrodis andA .placenta cause dove r90 %mortalit yi nwhitefly , whileV. lecanii caused 50% mortality. On poinsettia, whitefly mortality was significantly lower for all three fungi (ca. 20%). The percentage of aphidsinfecte d byV. lecanii orMetarhizium anisopliae wassignificantl y higher on gerbera than on cucumber. With V.lecanii thehighes t percentage mycosis was obtained. The phyllosphere of cucumber was more humid than that of the other crops. This cannot explainhighe raphi dmycosi so ngerber acompare d tocucumber ,no rlowe rwhitefl y mortality onpoinsettia .Th efac t thatpoinsetti aleave swer emor ehydrophobi c thanth eothe rleave sma y offer anexplanatio nfo rth eobserve dlowe rwhitefl y mortality,bu tchemica lhost-plan taspect s mayals opla ya role .Ou rresult sunderlin e theimportanc e of thefirs t trophic level (plant) for entomopathogenic fungi inintegrate d pestmanagemen t programs. Theinfluenc e ofrelativ ehumidit y(RH )an dhost-plan tspecie san dthei rinteractio no n mycoses of T. vaporariorum by entomopathogenic fungi A. aleyrodis, A. placenta and V. lecanii was studied in chapter 7. Experiments conducted on poinsettia, gerbera and cucumbera t50 %an d 80%R Hshowe da clea rhost-plan teffect . Oncucumbe ran dgerber a all fungi caused significantly higher mortality compared with the control, whereas whitefly mortality on poinsettia remained low and was not significantly different from the control (0.05% Tween). A higher RH resulted in a higher mortality caused by the fungi. Both Aschersonia spp.performe d significantly better than V.lecanii. I nfollowin g experiments an 165 Summary additionalperio do f0,3,6 , 12,2 4o r4 8hour so fhig hR Hwa sapplie d after treatment withA. aleyrodis andV. lecanii.Mortalit y caused by the fungus increased with a longer additional period of high RH and higher ambient humidity. However, the host-plant species effect exceeded the effect of ambient or additional high RHperiod ;whitefl y mortality was highest on cucumber and lowest on poinsettia. On poinsettia both fungi had hardly any effect on whitefly mortality,excep t after a4 8hr shig hR Hperiod . Ongerber a andcucumbe r all fungal treatmentswer esignificantl y different from thecontro lan dA. aleyrodisperforme d bettertha n V. lecanii. The influence of cultivar differences on the efficacy of entomopathogenic fungi is testedi nchapte r8 .Th eexperimen t wasconducte d ina glasshous e (semi-practice) using two gerberacultivars , 'Bourgogne' and 'Bianca' withdifferen t plantstructur ean dtrichom edensit y on the leaf: 'Bianca' having more and larger leaves and fewer trichomes/cm2 than 'Bourgogne'. Two entomopathogenic fungi, A. aleyrodisan dV. lecaniiwer eapplie d for the control of T. vaporariorum. The fungi were applied in two concentrations, 106 and 107 conidia/ml.I n 'Bianca', both fungi caused awhitefl y mortalityu pt o 80%. Whitefly mortality was higher for 107tha n for 106conidia/ml .I n 'Bourgogne', V.lecanii cause d a significantly lower mortality thanA aleyrodis.Althoug h no cultivar differences in whitefly development time were found, other characteristics, like natural mortality and build-up of the whitefly population, seemed to differ. Differences between results in mortality in relation to cultivar (humidity, hit-probability) and whitefly characteristics mayexplai n the differences between theefficac y ofth efunga l species. The host plant is the most important factor in this tritrophic system, exceeding the influence of humidity. However, although entomopathogenic fungi are less successful in controlling whiteflies on poinsettia, these problems may be circumvented by use of formulation. Aschersonia spp. are highly virulent against whitefly pests, can cause natural epizootics in the field, can be grown on artificial media and applied with conventional insecticide application equipment, arewel l adapted tosurviv ei nth ecanop yenvironment , and are compatible to/or even complementary with other natural enemies. 166 Samenvatting Witte vlieg:biologie ,plaag ,schad e en bestrijding Integenstellin gto twa td enaa mdoe tvermoede nzij n wittevliege ngee nvliegen ,maa rbehore n zeto td esapzuigend einsecte n (Homoptera),waarto eoo kd ebladluize nbehoren .E r zijn meer dan 1200witt e vlieg soorten beschreven, maar slechts enkele soorten vormen een probleem binnen de landbouw. In kasteelten zijn twee soorten schadelijk, nl. de kaswittevlieg, Trialeurodes vaporariorum,e n de tabakswittevlieg, Bemisiaargentifolii. D e kaswittevlieg vormta l enkele decenia een probleem, detabakswittevlie g ispe rongelu k gei'ntroduceerdi n 1987. Wittevlie gheef t gevleugelde adulten.D evrouwtje s leggen huneiere nbi j voorkeuro p deonderzijd e vanjong ebladeren .He teerst elarval estadiu mi sno gmobiel ,maa rzodr ahe tee n geschikte voedingsplek op een blad gevonden heeft, zal de larve tijdens zijn hele larvale stadium opdezelfd e plekblijven . Delarv esteek t z'n styleti n hetbla d omzic h tevoede n met plantensappen enza ldez eallee nterugtrekke n tijdens devervellin g naarhe tvolgend e stadium. Witte vlieg heeft vier larvale stadia die steeds in grootte toenemen. Het vierde stadium ontwikkelt zich tot pop, wat betekent dat er geen afzonderlijk pop-stadium is, maar dat de adult zichbinne n dehui dva nhe tvierd estadiu m ontwikkelt.Uiteindelij k zal denieuw e adult door scheuren in het exoskelet naar buiten komen. De duur van deze levenscyclus is afhankelijk van deplantensoor t end e temperatuur. Witte vlieg kan op verschillende manieren schade veroorzaken aan gewassen. Ten eerste, als witte vliegen zich massaal aan een plant voeden, kan de plant zich minder goed ontwikkelen, wat kan leiden tot opbrengstderving, het laten vallen van bladeren, etc. Ten tweede scheidt dewitt evlie gd eovermaa t aan suiker weer uit, dezogenaamd e honingdauw. Dit stimuleert degroe i vanroetdauwschimmels ,wa t deplan t kan remmen, maar belangrijker isd eeconomisch eschade :vruchten ,bloeme ne nkatoenplui sworde nbedek tme tee nplakkeri g laagje.Te nderd eka nwitt evlie gverschillend eplantenvirusse n overbrengen. Eenenkel ewitt e vlieg besmet met een plantenvirus kan grote problemen veroorzaken binnen een teelt. Ten vierde gelden voor export van planten of plantenproducten vaak nultoleranties, er mag dan geen enkele witte vlieg op het product aanwezig zijn. Als de wittevliegsoort in een importerend landee nquarantain einsec t is,ka ndi tleide n totvernietigin gva nd etotal elading . In natuurlijke ecosystemen en landbouwsystemen waar geen pesticiden gebruikt worden vormt witte vlieg een veel minder groot probleem, omdat ze op een laag niveau gehouden worden door verschillende natuurlijke vijanden: predatoren, parasitoiden en pathogenen.Predatore n dodenhu nproo idoo rdez eo p teete n of leegt e zuigen. Parasitoiden, 167 Samenvatting bv. sluipwespen, leggen eieren in het insect (gastheer) en de ontwikkelende parasitoidlarve consumeert het insect tot dat het sterft. Pathogenen zijn ziekteverwekkers zoals bacterien, schimmels en virussen, die vaak hun gastheer doden. Door het gebruik van bestrijdingsmiddelen zijn deze natuurlijke vijanden vaak uitgeroeid en kan witte vlieg een plaag worden.Verde rhebbe n verandering in gewasrotatie, verkorting vanbraakperiode s en opeenvolgende teelten vanwittevlieggevoelig egewasse nerto egelei d datnatuurlijk e vijanden nietlange r in staat waren witte vlieg opee n voldoende laagnivea u tehouden . Chemische bestrijdingsmiddelen zijn nad eTweed e Wereld Oorlog de belangrijkste methodegeworde n voord ebestrijdin g vaninsectenplagen . Zeware n eenvoudig toet epassen , vormden een goede bescherming tegen plagen en waren betrouwbaar. Pas later werden de nadeligegevolge nduidelijk : risico'svoo rmen se nmilie ue nresistenti eva ninsecte ntege ndez e middelen. Ditga f aanzetto t onderzoek naar anderebestrijdingsmethodes . Het huidige beleid is erop gericht de afhankelijkheid en het gebruik van chemische bestrijdingsmiddelen terug te brengen. Een belangrijk alternatief is biologische bestrijding waarbij predatoren, parasitoiden enpathogene n worden ingezet omee n plaag onder controle te krijgen. Pathogenen In ditproefschrif t staan insectpathogene schimmelste rbestrijdin g vanwitt evlie gcentraal .Z e werkenrelatie f snelt.o.v .predatore n enparasitoide n enzij n metconventionel e spuitappartuur toet edienen .He tinfecti eproce szie te ral svolg tuit :ee nspor edi eo phe tinsec tbeland tkiemt , penetreert decuticul a enbetrek t zijn voedsel uit het het insect. Het insect zalhieraa n sterven end eschimme lgroei tvervolgen s weernaa rbuite n omt e sporuleren. Na verspreiding van de sporen door bv. wind of water zal hetproce sopnieu w beginnen. Dezeschimmel skunne nonderverdeel d wordeni ntwe egroepen ,specifiek e schimmels dieallee n wittevlie ginfecteren , zoalsAschersonia-soorten e nbree d werkendeschimmels ,di e insecten uit verschillende ordes kunnen infecteren en soms ook andere schimmels, voorbeelden hiervan zijn Beauveria bassiana, Paecilomycesfumosoroseus an d Verticillium lecanii. Het voordeel vanhe t gebruik van specifieke schimmels isda the t andere natuurlijke vijanden niet zal infecteren, zezulle n echter ook andereplaaginsecte n niet infecteren. Verschillendecriteri akunne n wordengebruik tvoo rhe tselectere n vanee n succesvolle schimmel: gastheerspecificiteit, virulentie, persistentie, het gemak waarmee sporen (de infectieuse eenheid van schimmels) kunnen worden geproduceerd, etc.. Schimmels worden vaak op dezelfde manier toegepast als chemische bestrijdingsmiddelen. De sporen die door schimmels worden geproduceerd worden op het aangetaste gewas gespoten, vaak in hoge doses en de behandeling wordt enkele keren herhaald, vandaar dat er gesproken wordt over myco(=schimmel)pesticiden. Selectiecriteria alsvirulentie , nl.ee n zohoo gmogelijk e sterfte 168 Samenvatting bereikenme tee nz olaa gmogelijk edosi si nee nz okor tmogelijk e tijd, enproductiegema k zijn erg belangrijk. Een schimmelsoort kan bestaan uit genetisch zeer diverse isolaten/stammen, die aanzienlijk kunnen verschillen in virulentie t.o.v. een plaag. De virulentie kan door verschillende factoren worden bei'nvloed. Zo kan een schimmel die vaak op kunstmatig mediumi sgekweek t zijn virulentie verliezen. Ook omgevingsfactoren kunnen de efficientie van een schimmel bei'nvloeden. Over het algemeen hebben schimmelsporen een zeer hoge relatieve luchtvochtigheid nodig om te kunnen kiemen (meer dan 92%), wat lang een bottleneck isgewees tvoo rhe tgebrui kva n schimmelsvoo rbiologisch ebestrijding . Daarnaast kan het insect zelf en/of de plant waarop het insect voorkomt, de waardplant, een grote rol spelen. Aschersonia Schimmelsoorten uit het geslacht Aschersonia infecteren alleen wittevliegsoorten en/of dopluissoorten.D eschimmel szij n afkomstig uit(sub)tropisc hgebied ,zij n vaakopvallen dva n kleur,z ekleure n bv.hu n gastheer oranje, enzij n in staathu n gastheer volledig uit teroeien . SommigeAschersonia-soorten zij n specifiek voorwitt evheg ,ander esoorte ninfectere n alleen dopluis. Doordat de larvale stadia van dopluis en witte vlieg sterk op elkaar lijken, is er discussieove rd especificitei t vand eschimmel ,omda td egastheersoor t(witt evlie go f dopluis) nainfecti e vaaknie t meer teherkenne n is. VerschillendeAschersonia soortenwerde na laa nhe tbegi nva nd evorig eeeu wingeze t voorbiologisch ebestrijdin gva ncitruswittevlie gi nFlorida .D eschimme lkom tdaa rva nnatur e voor, maar werd kunstmatig gekweekt zodat citrustelers de schimmel in overmaat in hun boomgaarden konden inzetten. Ind ejare n 60e n7 0 werden verschillende soorten ingezet ter bestrijding van citruswittevlieg en kaswittevlieg in verschillende Oost Europese landen. Doordat deze schimmels zeer specifiek zijn, dus te gebruiken naast andere natuurlijke vijanden, zei nstaa tzij n lango pbladoppervlakte s teoverleve ne nminde r afhankelijk zijn van hogeluchtvochtigheid , ise rno gsteed sinteress e voordez eschimmels . Doele nopbou w vanhe t proefschrift Het doel van dit onderzoek was tweeledig. Ten eerste, het selecteren van een virulente Aschersonia stamdi ezowe ltabaks -al skaswittevlie g kanonderdrukke n (hoofdstuk 2t/ m4) . Ten tweede, het bepalen van factoren die de efficientie van entomopathogene schimmels bei'nvloeden, inhe tbijzonde r deinvloe dva nluchtvochtighei d enwaardplan t (hoofdstuk 5t/ m 8). 169 Samenvatting In hoofdstuk 2 worden verschillende Aschersonia isolaten getest op hun vermogen goed te sporuleren op kunstmatig medium en hun virulentie tegen kas-e n tabakswittevlieg. Hieruit worden 6 isolaten geselecteerd die in hoofdstuk 3 nader bekeken worden, gezocht wordt naaree nsta mdi eme tee n zolaa gmogelijk e dosisi nee n zokor t mogelijke tijd in staat isbeid ewittevliegsoorte n teonderdrukken . Inhoofdstu k 4word the tinfectieproce s vanenkel e isolaten nader beschreven aan de hand van scanningelectronenmicroscoopfoto's en gevoeligheid van de verschillende wittevlieg stadia. In hoofdstuk 5 is de persistentie van A. aleyrodis onderzocht op verschillende gewassen. Of de gevonden verschillen tussen de gewassen gerelateerd kunnen worden aan de luchtvochtigheid opbladnivea u vlak rond het blad is onderzocht in hoofdstuk 6. In hoofdstuk 7 wordt de invloed van de omgevingsluchtvochtigheid enhe t gewas opd eeffectivitei t van insectpathogene schimmels beschreven.Uiteindelij k zijn 2schimmel si nhoofdstu k 8o pkasnivea u metelkaa r vergeleken, waarbij tevens de invloed van rasverschillen onderzocht is. In hoofdstuk 9 worden de belangrijkste resultaten in het kader gezet van biologische bestrijding met behulp van schimmels. Hoofdstuk 2: Vierenveertigisolate nva nverschillend eAschersonia-soorten zij n getesto phu nvermoge no m goed te sporuleren op kunstmatig medium en hun virulentie tegen kas-e n tabakswittevlieg, beidebelangrijk e voorwaarden voorhe t succesvol gebruik van schimmels ter bestrijding van witte vlieg.D e4 4isolate n waren afkomstig uit verschillende schimmelcollecties, daarnaast zijn enkele stammen door de onderzoekster zelf gei'soleerd uit geinfecteerd wittevliegmateriaal. De isolaten komen oorspronkelijk uit (sub)tropische gebieden van Amerika (25isolaten) ,Azi e(1 7isolaten ) enAfrik a (2isolaten) . Metuitzonderin g vanee nisolaat , sporuleerden alleisolate n in meero f minderemat e opgeautoclveerd e gierst,hoewe lsommig eisolate n beter sporuleerden opmaismedium .Ach t isolaten sporuleerden zeer goed met ruim meer dan 5 x 10' sporen per culture. Kiemingspercentages opwateraga rverschilde n sterk,varieren d tussen 3.4 en 99.5%, waarbij enkeleisolate nnaas tgewon ekiembuize noo kcapilliconidi a produceerden. Capillicondia zijn conidia die gevormd worden op een haarachtige structuur. Bij andere entomopathogene schimmelsdiene ndez estucture n voorhe tinfectere n vanpasserend einsecten .Welk e functie deze sporen bijAschersonia-soorten vervullen is nognie t bekend. Van desporulerend e isolaten waren ernege n niet in staat kas-e ntabakswittevlie g te infecteren, ofkonde nnie topnieu wgei'soleer dworde nui td einsecten .Di tka nt emake nhebbe n met de specificiteit van deisolaten . Als de oorspronkelijke gastheer nietbeken d is,kunne n insolaten van dopluis afkomstig zijn. Deze isolaten zijn waarschijnlijk niet in staat om witte 170 Samenvatting vliegt einfecteren . Hetka noo kzij nda td eisolate na lz olan go pkunstmati gmediu mgekweek t zijn dat zehe t vermogen omwitt evlie g teinfectere n kwijt zijn geraakt. Deherisolate n zijn getoetst ophu nvirulenti etege nbeid ewittevliegsoorten . Meerdere isolaten waren in staat een hogesterft e onderbeid ewittevliegsoorte n teveroorzaken . Virulentie tegenkaswittevlie g was danoo kpositie f gecorreleerd metvirulenti e tegen tabakswittevlieg.E rwa sechte rgee n relatie tussengoe dkiemend ee nvirulent eisolaten .Uiteindelij k zijn zesisolate ngekoze no pbasi sva n sporulatie (A31, Aa5 en Apl), virulentie (A24, A26, A31, Aa5 en Apl) en geografische oorsprong (A23). Hoofdstuk 3 Deze zes isolaten zijn vergeleken op basis van hun LD50: de dosis waarbij 50% van de wittevliegpopulatie overlijdt, en hun LT50: de tijd die nodig is om 50% van de populatie te doden.Ho elage rd eLD 50e nho ekorte rd eLT 50ho evirulente rhe tisolaa tza lzijn . OmdatA2 3 vaak niet in staat was meer dan 50% van de populatie te doden zijn de LD50 en LT50 niet berekend voor dit isolaat. De LD30 varieerde voor tabakswittevlieg van 16 (A26) tot 48 sporen/mm2(Aa5 )e nvoo rkaswittevlie g van7 (A24)to t 53sporen/mm 2 (Aa5),dez e waarden waren echter niet significant verschillend van elkaar. Bij de hoogste toegepaste dosis (108 sporen/ml) laghe t infectiepercentage voor isolaten A23e n A24 lager in vergelijking met de op een na hoogste dosis (107 sporen/ml). Deze zelfinhibitie is een fenomeen dat vaker geconstateerd isbi j schimmelse ndien to mkiemin gen masse tevoorkomen .D edosi s waarbij dit fenomeen optreedt is echter zo hoog dat het geen weerslag zal hebben op biologische bestrijding vanplagen . De LT50 kan bij niet-mobiele insecten alleen bepaald worden aan de hand van uitwendigeinfectiesymptomen , inhe tgeva lva nAschersonia doo rhe tverkleure nva nd elarve n van doorzichtig naar parelmoerachtig oranje (A24, A26, Aa5 of Apl), wit (A23) of bruin- oranje (A31).Voo rtabakswittevlie g varieerded eLT 50va n5 (A31 )to t 8dage n (A24) envoo r kaswittevliegva n5 (A31/A26 )to t 10dage n(Apl) .He tgbrui kva nuitwendig esymptome nka n leiden totee nonderschattin g vand eLT 50,omda t sterfte vaaka loptreed t voordat symptomen zichtbaar worden. Er isdaaro moo kee n tweede methode gebruikt, nl.he t tijdstip van sterfte aand ehan d vanhe tlarvestadiu m waarin sterfte plaatsvindt bepalen. Deze gegevenskwame n globaal overeen met de gegevens van de LT50 waardes, echter isolaat Apl overtrof hierbij andereisolate n i.t.t.d eLT 50waarden . Het verkleuren van delarve n nainfecti e door Apl zou weleen s trager kunnen verlopen dan bij de andere isolaten. Uithoofdstu k 2e n3 blijk t datisolate n A26,A31 , Aa5e nAp l goedekandidate n zijn voorbiologisch ebestrijdin g vanwitt evlieg . 171 Samenvatting Hoofdstuk 4 Het infectieproces van tabakswittevlieg door A. aleyrodis (Aa5), A. placenta (Apl) en Aschersonia sp. A26, is in meer detail bekeken d.m.v. scanning electronen microscopic Daarnaast is de gevoeligheid van de verschillende stadia van kas- en tabakswittevlieg voor Apl enAa 5 onderzocht. Uit bovenstaande waarnemingen bleek dat dedri e isolaten nauwelijks kiemen opd e bladschijf van kerstster (poinsettia), maar wel op de bladnerven en op de insecten zelf. De sporen ophe t insect produceerden een slijmlaag waardoor zezic hbete r kunnen hechten aan het insect. Voordat de kiembuizen de huid van het insect penetreerden werden vaak appressoria gevormd. Dit zijn structuren waarmee de schimmel zich extra goed aan de ondergrond kan hechten, watuiteindelij k penetratie vergemakkelijkt. Echter, in even zoveel gevallen bleef bij penetratie de vorming van appressoria achterwege.D epenetratieplaat s en heta lda nnie tvorme nva nappressori alee knie tgerelateer d aanee nspecifiek e pleko pd ehui d vanhe tinsect .Som swerde ndoo rd eschimmel see naanzienhjk e hoeveelheid schimmeldraden op het insect gevormd zonder dat infectie plaats leek te vinden. Eerste tot en met derde stadium larven van zowel kas- als tabakswittevlieg zijn zeer vatbaar voor infectie van A. aleyrodise nA. placenta, vierde stadium larven enpoppe n zijn veelminde r vatbaar. Deze verminderde vatbaarheid heeft waarschijnlijk te maken met het ontwikkelen van de adulte witte vlieg binnen de huid van de vierde stadium larve. Voor beide wittevliegsoorten is er opmerkelijk veel overeenkomst in vatbaarheid van de verschillende larvestadia. In het algemeenvon d sterfte plaatsi n een stadium later dan het behandelde stadium, d.w.z. dat80 - 90% van depopulati e binnen 5 a7 dagen stierf. Onder gunstige omstandigheden, leeshog e luchtvochtigheid, komt de schimmel via dezwakk e plekken van het insect weernaa r buiten omt e sporuleren. Hoofdstuk 5 Vooree n succesvolle infectie vanwitt e vlieg zalee n sporei n contact moetenkome nme t het insect.Echte rbi jbespuitinge n vanplante nzulle nmaa rweini g sporendirec ti n contactkome n met het insect. Doordat het eerste stadium van witte vlieg beweeglijk is en doordat bij opeenvolgende vervellingen delarve n steedsi n groottetoeneme nkunne n deze sporen alsnog in contact komen met hun gastheer. Hiervoor moeten zeechte r wel langetij d levensvatbaar blijven .O fspore nn aee nlan gverblij f ophe tbla dno gi nstaa tzij nt ekieme no pwateraga r(ee n maatvoo rlevensvatbaarheid) , daarnaast in staat zijn hun gastheer teinfectere n en wat dero l vand eplan ti ndez eis ,i si n dithoofdstu k onderzocht. Dekiemkrach t van sporen vanA. aleyrodisblee f erglan gbehouden , nl. na 31dage n opee nkomkommerbla d bleek 80a 90 %va n de sporenno gsteed si n staato mt ekiemen ; op 172 Samenvatting kerststerbladblee kdi tgemiddel d70 %t ezij ne no pgerberabla d 'slechts'gemiddel d50% .Voo r komkommere ngerber ablee kdi tee ngoed emaa tvoo rd einfectiekan s vankaswittevlieglarven , nl. 31dage nn abespuitin g werd opkomkomme r nog steeds 90%va n delarve n gedood door infectie, opgerber a varieerde het dodingpercentage tussen 50e n 70%. Opkerstste r bleef het dodingspercentage ver achter bij de verwachting, het zakte nl. van 70% op de dag van bespuiting naar 20% op dag 3 tot en met dag 31 na bespuiting. Op bladeren van de drie gewassenblee ke rnauwlijk s kiemingt ezij n opgetreden,wa tka nwijze n opee n"zi te nwacht " strategicva nA. aleyrodis sporen inplaat s van tekieme ne nnaa r hun gastheer toe te groeien. Dat de plantensoort de kiemkracht en infectiekans sterk bei'nvloedt, kan te maken hebben met verschillende factoren. Teneerste ,d esamenstellin g van microorganismen ophe t bladverschil t sterkva nplan tto tplant ,dez ekunne n inwerken opinsectpathogen e schimmels viabv .competiti e omvoedingsstoffe n ophe tblad ,etc. .Z oblijk t dat zowelkomkommer - als gerberablad aangetast kunnenworde ndoo rverscheiden e schimmelziektes,terwij l bij kerstster slechtsee nbladpathogee n bekend is.Te ntweede ,d echemisch e samenstelling van deplante n verschilt aanzienlijk, wat kan leiden tot directeremmin gdoo r deplan t danwelbei'nvloedin g van de vatbaarheid van het insect. Zo is van kerstster bekend dat deze plant giftige stoffen bevat die celdodend kunnen werken. Ten derde, de planten verschillen aanzienlijk in morfologie, bv.komkommerbladere n zijn sterkbehaar d en aanzienlijk groter in vergelijking metbladere n vankerstster .Dez emorfologisch e kenmerken kunnen deluchtvochtighei d in de laag op het blad, waarin het infectieproces zich afspeelt, bei'nvloeden. Ondanks de grote invloed van deplantensoor t opd epersistenti e vanA. aleyrodisi sd elang eperiod e dat deze schimmeli n staat iskaswittevlie g te infecteren zeer opmerkelijk. Hoofdstuk 6 De infectieprocessen vinden plaats in een dunne laag op hetbla d en de luchtvochtigheid in deze laag is van groot belang voor de effectiviteit van insectpathogene schimmels. Deze fyllosfeerluchtvochtigheid wordtbepaal d door omgevingstemperatuur, -luchtvochtigheid en wind,maa roo k dooreigenschappe n van deplant , zoalsbladvor me n -grootte, bladbeharing, dichtheid van het gewas, etc. De fyllosfeerluchtvochtigheid kan echter niet direct gemeten worden zonder deze tebei'nvloede n en kan alleen benaderd worden door ander factoren als omgevingstemperatuur, bladtemperatuur e.d. temete ne n vervolgens de luchtvochtigheid te berekenen.D efyllosfeerluchtvochtighei d isbereken d voorgerbera ,komkommer , kerstster en tomaat, waarbij kaswittevlieg op alle gewassen was uitgezet, terwijl katoenluis alleen op komkommer engerber a wasuitgezet .D efyllosfeerluchtvochtighei d isvervolgen s gerelateerd aan de effectiviteit van verschillende insectpathogene schimmels op deze gewassen. De schimmelsA. aleyrodis, A. placentae n Verticilliumlecanii zij n gebruikttege n wittevlieg ,e n 173 Samenvatting Metarhizium anisopliaee nV. lecaniizij n gebruikt tegenkatoenluis . Defyllosfeerluchtvochtigheid o pkomkomme r wassignifican t hogerda n opd e andere gewassen en lag gemiddeld rond de 94%.D e meeste schimmelsporenhebbe n een minimale luchtvochtigheid van9 2a 93 % nodigo mt ekiemen .Beid eAschersonia soorten veroorzaakten meer dan 90%sterfte , V.lecanii slecht s 44%.Ondank s de lagere fyllosfeerluchtvochtigheid optomaat -e ngerberabla d (90%),veroorzaakte noo khie rbeid eAschersonia soortenee n sterfte van90 %o fmeer ,terwij lV. lecaniihe t minder goeddeed .Hoewe l schimmels in staat zijn om tekieme nbi j 92%,gaa the t snelleral sd eluchtvochtighei d hogeris ,wa tka n betekenen dat de berekendeluchtvochtighei d hieree nonderschattin g is vand ewerkelijk e luchtvochtigheid op bladniveau. Hoewelo pkerststerbla d deluchtvochtighei d ophetzelfd e niveaula gal so pgerber ae n tomaat,blee khie rgee n van deschimmel s in staat een hogewittevliegsterft e te veroorzaken, nl.21% ,wa tnie tverschild eva nd esterft e ind econtrolebehandeling .D ebladere nva nkerstste r zijn sterkwaterafstoten d invergelijkin g metd eander eplanten ,wa tgevolge nka nhebbe n voor de verdeling van de sporen over het bladoppervlak. Dat de fyllosfeerluchtvochtigheid de gevonden verschillen in sterfte nietka n verklaren wordtbevestig d doord ebladluisgegevens ; debladluissterft e veroorzaakt door schimmels laghoge r op gerbera dan op komkommer. Naastd efysisch e eigenschappen vand eplante nkunne noo kchemisch e eigenschappen een rol gespeeld hebben in de verschillen. Zo kunnen toxische stoffen van de plant de schimmel direct bei'nvloeden, maar ook insecten beschermen tegen schimmelinfecties. Het verschil tussen deschimmel ska nveroorzaak t zijn dooree nverschi l in 'droogte'tolerantie .Z o heeft V. lecaniiee n hogere luchtvochtigheid nodig dan A. aleyrodis.Echter , het verschil in mortaliteit tussen detwe eschimmel ska n ook veroozaakt zijn dooree n verschil in virulentie. Hoofdstuk 7 De rol van de omgevingsluchtvochtigheid, plantensoort en hun interactie in het succes van schimmelinfecties is hier nader onderzocht. Komkommer- gerbera- en kerststerplanten gei'nfesteerd metkaswittevlie g werden behandeld metA. aleyrodis,A. placenta enV. lecanii onder droge en vochtige omstandigheden al dan niet met een extra periode van zeer hoge luchtvochtigheid (>95% ) varierend van 3t o4 8 uur. Dit laatste werd alleen metA. aleyrodis enV. lecaniiuitgevoerd . Over het algemeen bleek dat een hogere omgevingsluchtvochtigheid leidde tot een hogere infectie, echter metuitzonderin g van kerstster. Dat deplantensoor t van groot belang is blijkt wel uit het feit dat op kerstster de sterfte veroorzaakt door de schimmels alleen verschiltva nd econtrol ewannee rd eperiod eme textr ahog eluchtvochtighei d4 8uu ro flange r is, terwijl op komkommer bij een omgevingsluchtvochtigheid van gemiddeld 45% zonder 174 Samenvatting extraperiod ehog eluchtvochtighei dA. aleyrodisa l50 % sterfte veroorzaakte.Verde rblee kda t A. alyerodis(e nA. placenta)effectieve r zijn bij lageluchtvochtighede n danV. lecanii. Uit deze gegevens blijkt dat wanneer de omgevingsluchtvochtigheid maar gunstig genoeg is, de grote verschillen die gevonden zijn in effectiviteit van de schimmels op de verschillende plantensoorten, verminderen. Door schimmels te formuleren tot een product, waardoor ze minder afhankelijk worden van luchtvochtigheid, kan dit probleem omzeild worden.Omgevingsluchtvochtighei d beinvloedt het infectieproces enka n evt. verschillen in effectiviteit van schimmelsbinne nee nplantensoor t verklaren,maa rd eplantensoor t overtreft het effect van omgevingsluchtvochtigheid. Hoofdstuk 8 Twee schimmels, nl.A. aleyrodise n V.lecanii zij n toegepast onder semipraktijkomstandig- heden inee ngerberagewa s metee ngemiddeld e omgevingsluchtvochtigheid van 70%.E r zijn vele gerberarassen op de markt (> 600) en deze kunnen aanzienlijk verschillen in morfologische danwelchemisch e eigenschappen. Indi thoofdstu k isgewerk t mettwe e rassen die gekozen zijn opbasi s van hun morfologische verschillen. 'Bianca' vormt een zeer dicht gewasme tgemiddel d8-1 5 relatief grotebladere npe rplan te nee nlicht ebladbeharing , terwijl 'Bourgogne'ee n relatief open gewas vormt met gemiddeld 5-7 relatief kleinere bladeren per plantme tee nzee rdicht ebladbeharing .Naas teffectivitei t vand eschimmels ,di etoegepas t zijn intwe edoses ,t ewete n 106e n 107spore npe rml ,i ndri eopeenvolgend eweken ,i soo kgekeke n naard eontwikkelin g van kaswittevlieg opbeid e rassen. Uit het kasexperiment blijkt dat hoe hoger de toegepaste dosis hoe hoger de wittevliegsterfte door schimmelinfecties. Op 'Bianca' veroorzaken zowel A. aleyrodis als V.lecanii eenhog e wittevliegsterfte, nl. 80%. Datbeid e schimmels even goedpresteren , dit integenstellin gto tvoorgaand eresultaten ,ka nt emake nhebbe nme the tmeervoudi gtoepasse n van de schimmel in dit experiment. Op 'Bourgogne' echter, doet V. lecaniihe t significant slechter in vergelijking met A. aleyrodis en in vergelijking met 'Bianca'. Hiervoor zijn verschillende verklaringen aan tevoeren . Teneerste ,ee n sterkbehaar d blad ('Bourgogne') ismee rwaterafstoten d dan een licht behaard blad ('Bianca'). In het laatste geval vloeien de druppels verder uit, wat de luchtvochtigheidpositie fka nbeinvloeden .Same nme td edichter estructuu rva nhe tgewa ska n de luchtvochtigheid op het 'Bianca'blad hoger zijn dan op 'Bougogne'. Gezien A. aleyrodis minder afhankelijk is van hoge luchtvochtigheid kan dit de hogere sterfte, in vergelijk met V.lecanii, verklaren . Ten tweede, door de slechtere verdeling van sporen op 'Bourgogne'-blad (waterafstotendheid v/h blad) en doordat wittevlieglarven beter beschermd worden door de beharing,za ld eraakkan so p 'Bourgogne'kleine rzij n dano p 'Bianca'.Doorda tA. aleyrodisi n 175 Samenvatting staat is langer te overleven op een blad dan V. lecanii, zou de infectie van A. aleyrodis uiteindelijk hoger kunnen zijn. Tenderde ,hoewe l dewittevliegontwikkelin g opbeid erasse n nietverschilt , hebbend e adultewitt evliege nee nvoorkeu rvoo r'Bianca' .O mtoc hee nvergelijkbar e wittevliegpopulatie op'Bourgogne 't ecreeren ,i se rvake ruitgeze t in 'Bourgogne'da n in 'Bianca'.Di theef t geleid tot een 'jongere' populatieopbouw in 'Bourgogne'. Op het moment van bespuiting waren er waarschijnlijk meer eieren op 'Bourgogne' aanwezig en gezien A. aleyrodislange r overleeft opbladoppervlaktes , kan deze schimmelhierva n geprofiteerd hebben. Uitdi texperimen t komtduidelij k naarvore n dat insectpathogene schimmels in staat zijn een wittevliegpopulatie in gerbera teonderdrukken , waarbijA. aleyrodisbete r uitgerust lijkt tezij n danV. lecaniio monde r wisselende omstandigheden goedt e functioneren. Hoofdstuk 9 Voorhe t gebruik vanAschersonia-sooften ind epraktij k zulleneers tno genkel e hindernissen genomen moeten worden. Voor grootschalig gebruik van schimmels is een effieciente massakweek nodig. Iedere schimmel heeft zo zijn specifieke eisen en hiervoor zal meer onderzoek noodzakelijk zijn. Nieuwe systemen die delaatst ejare n ontwikkeld zijn voorhe t effectief kweken van andereschimmel s zouden hierbij uitkomstkunne n bieden.Daarnaas t is voor gebruik in depraktij k een product nodig dat een half tot anderhalfjaa r houdbaar moet zijn, zonder z'n effectiviteit te verliezen. Het formuleren van een schimmel tot een product heeft voordelen, zo kan bv. de schimmel beter beschermd worden tegen ongunstige omstandigheden, bovendien kan het gebruiksgemak verbeterd worden. Echter het infectieproces magnie t gehinderd worden doord e formulering. Deinteress e voorhe tgebrui k vanbiologisch ebestrijdin g inhe t algemeeni sgroeiend e vanwegeresistentieontwikkelin g bij plaaginsecten tegen chemische bestrijdingsmiddelen en toenemende drukvanui td emaatschappi j vooree nmilieuvriendelijker eaanpa kva n ziektese n plagen. Uit bovenstaande resultaten en uit voorgaand onderzoek blijkt dat verschillende Aschersonia-soottenzee rvirulen tzij ntege nbeid ewittevliegsoorten ,kunne nworde ngekwee k opkunstmati g medium,zee rgoe dzij n aangepast aan hetbladsysteem , zeerpersisten t zijn en doorhu nhog emat eva nspecificitei t goedt ecombinere n zijn metander enatuurlijk e vijanden. Steeds minder nieuwe, selectievere chemische bestrijdingsmiddelen komen op de markt vanwegen hoge kosten voor ontwikkeling en registratie, en dit geldt nog sterker voor een kleinemark tal sd eglastuinbouw .Doorda tinsectpathogen eschimmel si nzoverr e vergelijkbaar zijn metchemisch ebestrijdingsmiddele n datz eo peenzelfd e manierkunne nworde ntoegepas t enrelatie f snelwerke n in vergelijking met bv.parasitoide n enpredatore n ise rmee r enmee r belangstelling voor deze categorie natuurlijke vijanden. 176 Nawoord Hetheef t heelwa tzweetdruppeltje s gekost,maa reindelij k ishe tzover :he tproefschrif t isaf! ! Het grootste deelva n het werk, dathie r wordtbeschreven ,i suitgevoer d op het Proefstation voor Bloemisterij en Glasgroente, in Aalsmeer (nu Praktijkonderzoek Plant en Omgeving, sector glastuinbouw), innauw e samenwerking met hetLaboratoriu m voor Entomologie van deWageninge n Universiteit.Hoewe lallee nmij nnaa mo pd ekaf t staat,wa sdi tonderzoe knie t mogelijk geweestzonde rd einze tva nvel emensen .Ee naanta lva nhe nwi li khierbi jme tnam e bedanken. Allereerst wili kJok eFranse n mijn copromotor enJoo p vanLentere n mijn promotor bedanken. Zonderjulli e was er geen project "insectpathogene schimmels tegen wittevlieg" geweest.Joke ,jou w enthousiasme voorhe tgebrui k vanAschersonia terbestrijdin g van witte vliegwa ser gaanstekelijk ;Aschersonia-soorten zij ninderdaa dintrigerend eschimmel so mme e tewerken .Bedank t voor dediscussies , correcties en deondersteunin g in deafgelope n jaren. Joop,bedank t voorj e steun enhe t corrigeren van allemanuscripten , desnelhei d waarmeej e dat deed hebi k altijd zeer gewaardeerd. Duizendmaal dank aan Nina Joosten, die twee jaar van het project met mij heeft samengewerkt. Nina, ik heb veel bewondering voor jouw enorme inzet bij de vaak arbeidsintensieve proeven.I khe bj e (bijna) nooit horen mopperen.Jou w enthousiasme heeft meove rmenig edi pheengeholpen .Elle n Beerling, onze projecten liepen parallel,jouw werk met insectpathogenen tegen bladluis en trips, mijn werk met insectpathogenen tegen witte vlieg.Di theef t uiteindelijk geleidto tee ngezamelij k artikel.Bedank tvoo r devel e discussies, het kritisch lezen van manuscripten en steun, het klankbord datj e samen met Joke en Nina vormde,i svoo rmi jer gbelangrij k geweest.Elle ne nNina , ikden kme tvee lplezie r terug aan delo ldi e wegeha dhebben . Ik zald eweddenschappe n tussenjulli e missen. Bijzondere dank aan Jan Tolsma, voorj e bereidheid ombi j te springen, het regelen achterd escherme ne nbijhoude n vand ewittevliegkweken . Ookwi li kMenn oBoogaart , mijn kamergenoot voor 4jaar , en de andere leden van de kerngroep gewasbescherming: Conny Lanser, Ineke Bosker, Albert Kerssies en Jan Amsing, bedanken die me met raad en daad hebbenbijgestaa n gedurendemij n Aalsmeer-dagen.All ecollega' s inAalsmeer , bedankt voor deprettig e werksfeer. Zonder de gewasverzorgers geen kasproeven; in Aalsmeer werden de kassen met plantenuitsteken dbijgehoude n doorPete rSchrama ,Han sSchiittler ,Jacque sva nZoest ,Joha n Meyvogel en Henk Koedijk, mannen, hartelijk dank. Verder wil ik Jacqueline Lamers (Aalsmeer),Wald od eBoe r (PlantResearc h International, WUR) en Gerrit Gort (Wiskunde, WU)bedanke n voorhet meedenke n enoplosse n van statistische vraagstukken. 177 Nawoord Auli Tapaninen, Sandor van Voorst en Jacco Schreurs hebben ieder voor een afstudeeronderwerp of stageenkel emaande n meegedraaid in het onderzoek. Jullie resultaten hebben bijgedragen tot een beter begrip van hetAschersonia - wittevliegsysteem , bedankt! Duringm ysecon dyea rI spen da wee ka tth eSchoo lo fBiologica l Siences (University ofBristol ,UK) .I wan tt othan kTari q Butt (IACRRothamsted/Univ . of Swansea, UK),Ala n Beckett andRober tPorte r (Univ.o f Bristol, UK) for this opportunity and their hospitality.I t waswonderfu l thatth e scanningelectro n microscope wasmad e available tome .Tariq , your enthusiasm iscontagious ! De begeleidingscommissie, bestaande uit: Gerrit Bollen (Lab. voor Fytopathologie, WU),Wille mRavensber ge nlate rRic kva nd ePa s(Kopper tBiologica l Systems),Ari e Strijbis (Poinsettiateler) ,Jaa pd eVrie s(Productscha pTuinbouw )e nEefj e denBelde r(Plan tResearc h International), heeft mij tijdens depraktisch e fase van het onderzoek bijgestaan met advies over verschillende onderwerpen, bedankt allemaal! Ook de car-poolers van het eerste uur, Annette Bulle, Ine van der Pluym en Martin Pieters, wil ik bedanken. Ik heb menig uurtje, op weg van Wageningen naar Aalsmeer en terug,me tjulli ei nd eaut odoorgebracht . Veleonderwerpe n werdenbesproke n enhe twa see n goedemanie ro mhe t werk in Aalsmeer achter telaten . Verder wili kmij n collega'so pentomlogi ebedanken ,me tnam ehe t "AIO-klasje"' dat heelwa tmanuscripte n vankritisch e opmerkingen heeft voorzien, enmij n (oud)collega' s van deBiocontrol groep opFytopatholgie .Allemaal ,bedank t voord e veleleuk euren ! En natuurlijk niet te vergeten mijn vrienden en familie. Marja, Mark, Thea, Kitty, Miriam, Ad, Rik, Eddy, Olga, Rene, Hans, Dine, Piet, Herman, Lindy, bedankt voor alle etentjes, wandelingen, kroegavonden, zonder "andere aandachtsvelden" was er nu geen proefschrift geweest.Pa pe nmam ,Ro be nAnita ,Ing ee nJuul ,bedank tvoo rjulli e vertrouwen, interesse en steun! Bedankt!! Wageningen, Mei 2001 Ellis 178 List of publications Meekes,E.T.M. , Fransen,J.J .& Lenteren ,J.C .van , 1994. Theus eo fentomopathogeni c fungi for thecontro l of whiteflies. Mededelingen vande Faculteit Landbouwwetenschappen RijksuniversiteitGent 59/2a ,371-377 . Meekes, E.T.M. & Fransen, J.J., 1995. Schimmel Aschersonia alternatief voor bestrijding wittevlieg. Vakbladvoor de Bloemisterij, nr.20 ,p.35 . Meekes,E.T.M. ,Franse nJ.J . &Lenteren ,J.C .van , 1996.Pathogenicit y of entomopathogenic fungi of the genusAschersonia agains t whitefly. BulletinlOBCwprs 19/1,103-106 . Meekes, E.T.M., Voorst, S. van, Joosten, N.N., Fransen, J.J. & Lenteren, J.C. van, 2000. Persistence ofth efunga l whitefly pathogenAschersonia aleyrodis, o n three different plant species.Mycological Research 104, 1234-1240. Meekes,E.T.M. , Beerling,E.A.M , Joosten, N.N. &Franse n J.J.. Role of host-plant climate in the interaction between entomopathogenic fungi and the two pest species Trialeurodes vaporariorum andAphis gossypii.Submitted . Meekes, E.T.M., Fransen J.J. & Lenteren, J.C. van. Virulence of Aschersonia spp. (Fungi Imperfecti) against whiteflies Bemisiaargentifolii and Trialeurodes vaporariorum. Submitted. Meekes,E.T.M. ,Joosten ,N.N. ,Fransen , J.J. &Lenteren , J.C. van.Effec t of gerbera cultivar on the efficacy of entomopathogenic fungi against Trialeurodes vaporariorum. Submitted Meekes,E.T.M. ,Joosten ,N.N. ,Franse nJ.J . &Lenteren ,J.C .van .Relativ ehumidit y andhos t plant species influence mycosis of whitefly, Trialeurodes vaporariorum, by Aschersoniaaleyrodis, A. placenta and Verticillium lecanii.Submitted . Meekes,E.T.M. ,Butt ,T.M. ,Joosten ,N.N. ,Porter , R., Beckett, A.,Fransen , J.J. &Lenteren , J.C. van. Germination and infection behaviour of Aschersonia spp. on Bemisia argentifoliian d Trialeurodes vaporariorum onpoinsettia . In preparation. Meekes, E.T.M., Fransen J.J. &Lenteren , J.C. van. Dose and time dependent mortality of whiteflies Bemisiaargentifolii an d Trialeurodes vaporariorum byAschersonia spp . (FungiImperfecti) . In preparation. Beerling, E.A.M. & E.T.M. Meekes, 1998. Biologische bestrijding van insecten met schimmels. Gewasbescherming 29(2),48 . Lenteren, J.C, Meekes, E., Yu Tong Qiu, 1999. Management of whitefies: new natural enemies andhost-plan t resistance.Bulletin IOBC wprs 22/1, 145-148. Obornik, M., Stouthamer, R., Meekes, E., & Schilthuitzen, M., 1999. Molecular characterization and phylogeny of the entomopathogenic fungus Aschersonia spp.. PlantProtection Science35,1-9 . 179 Curriculum Vitae Elisabeth (Ellis)Truu sMari aMeeke swer do p8 November 1967gebore nt eEibergen .I n 1986 behaalde zij hetVWO-diplom aaa nd e scholengemeenschap 'netMarianum 't e Groenlo en in datzelfde jaar begon zij met de studie plantenziektenkunde aan de Landbouwuniversiteit te Wageningen.Tijden s dedoctoraalfas e vandez estudi edee d zij onderzoek naard einvloe d van waarplantvarieteit, waardplantras engastheerdichthei d in het tritrofisch systeem kool -Pieris brassicae - Cotesia glomerata bij devakgroe p Entomologie ennaa r de herinplantingsziekte in asperge bij de vakgroep Fytopathologie. Een stageperiode werd doorgebracht bij Royal Sluis, San Juan Bautista/Salinas (Californie, USA),waa r zeonderzoe k deed naar methoden voor het toevoegen van microorganismen aan zaad ter bestrijding van omvalziekte(Pytium ultimum). In 1992behaald ezi j naardoctoraalexame n eni ndatzelfd ejaa r werktez ebi j Incotec (Enkhuizen). In 1994 werd ze aangesteld als AIO bij de vakgroep Entomologie in een samenwerkingsproject methe tProefstatio n voorBloemisteri j enGlasgroent e(Aalsmeer) ,ee n project gefinancieerd door het Productschap Tuinbouw. Dit werk heeft geleid tot het proefschrift getiteld: 'Entomopathogenic fungi against whiteflies: tritrophic interactions between Aschersonia species, Trialeurodes vaporariorum and Bemisia argentifolii, and glasshouse crops'. In 1999 - 2000 werkte ze als toegevoegd onderzoeker bij de vakgroep Fytopathologie aanwitt eroes t {AlbugoCandida) opkoo le nander ekruisbloemigen .Sind sme i 2001i sz ewerkzaa m alsecologisc h fytopatholoog bij PlantResearc h International. 181 Theresearc h described in this thesis was carried out atth eResearc h Station for Floriculture and Glasshouse Vegetables at Aalsmeer, the Netherlands in close cooperation with the Laboratory of Entomology of the Wageningen University. It is subsidized by the Product Board for Horticulture (Productschap Tuinbouw). On the front cover: -right : whitefly nymphs infected withAschersonia sp .A1 5 andA. placenta Apl - left: whitefly adult, healthy whitefly nymphs - top:poinsetti a crop and gerbera cv Bourgogne Cover design byMarjolei n deVette ,Beeldgroep ,WU R Printed byPonse n &Looije n BV, Wageningen