v.
Integrated pest manag
techniques and resource
Cotton growers generally rely on chemical treatments
to control serious infestations by a wide range of pests,
as immediate visible results can be obtained.
There are many risks associated with repeated pesticide
treatments, e.g. upsetting the established insect fauna
balance, development of pesticide resistance, possibility
of intoxicating farmers, and environmental pollution.
For several years, CIRAD entomologists have been
conducting studies in cotton fields with the aim of limiting
pesticide use through development of an integrated
approach to pest management involving various control
techniques.
C
otton crops are threatened by a very w id e range of pests, w ith a list of more than 70 a rth ro p o d pest species — m a in ly hom o p te ra ns (jassids, aphids and whiteflies), heteropterans (bugs and minds), lepidopterans (leaf- and boll- eating worms) and coleopterans, as w e ll as m ites (T a rs o n e m id a e and T e t r a n y c h id a e ) , d i p l o p o d s andM . VA ISSA YRE, J. C A U Q U IL , P. SILVIE n e m a to d e s (T a b le 1). In t r o p i c a l
C IR AD-C A BP 5 0 3 5 Africa, there can be losses of 30% to 3 4 0 3 2 M o n tp e llie r C e d e x 1, 100% of the crop yield potential due
_______________________ Frcmce to infestations of these pests.
Harvest debris in a cotton field.
Photo CIRAD-UREA
Chemical pest control is still essential in many agrosystems. Nevertheless, to reduce the dependence on pesti-cides for crop p ro tectio n, adapted cropping techniques, plant varieties w ith insect-tolerant traits, entomo- phagous insects, entomopathogenic agents and c h e m ic a l m e d ia to rs should be taken into serious conside-ration.
T his r e v ie w m a i n l y focuses on CIRAD projects under way in tropical Africa, as well as Latin America and Southeast Asia.
A larval Asopinae insect preying on a Spodoptera littoralis worm.
Photo CIRAD-UREA
ement
s
Cultural practices
This involves all crop management practices: so w in g, in te rc ro p p in g , planting density, weeding, m o n ito -ring plant growth and fertilization.
Plant and pest life cycles
C otton g ro w th and d e v e lo p m e n t cycles should be monitored in terms of the dynamics of the pest population present. In addipopulation, the flo w e -ring period, factors p ro m p tin g the fall of flowering and fruiting organs, and the plants' compensation poten-tial should be generally understood when making treatment decisions. In the African areas studied, plants sown e arly g e n e ra lly y ie ld e d the most cotton because of the fa v o u -rable climatic conditions. However, another type of crop is planted at the b e g in n in g o f the c ro p season, to avoid cotton pest attacks.
In central Côte d'Ivoire, before the advent of pyrethroids, farmers were thus advised to use the maize-cotton crop sequence to control C. leuco-
treta damage (ANGELINI, 1963).
Intercropping
In some countries, hardy cotton spe-cies (often Cossypium arboreum L.) are intercropped with other annual or perennial crops in traditional cotton cropping systems. This practice has c o n t i n u e d w it h the i n t r o d u c t i o n o f G. h ir s u t u m L., e s p e c ia l ly in S o u th e a s t A s ia (C R O Z A T
et a i , 1997). Although there have not been many investigations on the pest control impact, it seems that inter-c ropping various inter-crop speinter-cies w ith cotton can alter the population dyna-mics o f some pests (i.e. fa v o u rin g insects, or attracting them away from cotton plants), or boost ben eficial in s e c t p o p u la t io n s ( N I B O U C H E , 1995; SOGNIGBE, 1989).
Weed control
DEGUINE (1995) began an inventory of refuges of beneficiais for co ntrol-ling A. gossypii populations in nor-thern C am ero on . A lth o u g h weeds can shelter pest populations in crop- fields, they can also serve as a reser-v o ir o f b e n e fic ia l e nto m o ph a go us o rg a n is m s w h i c h c o u ld , in some cases, be managed to the benefit of the cotton crop (PERRIN, 1 975).
Fertilization
A few studies have highlighted that interactions between fertilization and crop p rotection techniques can be complementary, i.e. crop protection p ro g ra m m e s a da p te d to the c ro p potential have to be set up to enhan-ce the cost-effectiveness of fertilizer inputs (JOLY, 198 0; CRETENET & VAISSAYRE, 1986; EKUKOLE, 1992).
Destruction of harvest
debris
The traditional practice of destroying harvest debris can be very efficient when there is a well-defined interval between crop seasons. Pests adapt to the absence o f their host plants by undergoing diapause or migrating to o th e r hosts or sites. The red u c e d number of host plants during the dry season or the cold season reduces the su rv iv al p o te n tia l for species that cannot easily migrate.
Populations of monophagous or oli- gophagous insects that survive by undergoing diapause, especially the pink b o llw o rm (P. gossypiella), can be reduced by carefully destroying
Table 1. Inventory of insects referred to in the present article.
Family Genus and species
Coleoptera Anthonomus grandis (Boheman) Cheilomcnes sp.
Exochomus sp. Orthoptera Oecanthus sp. Homoptera Amrasca spp.
Amrasca biguttula (Ishida) Aphis gossypii (Clover)
Bemisia tabaci (Gennadius)
Jacobiasca spp.
Lepidoptera Alabama argillacea (Hübner) Amsacta meloneyi (Druce)
Anagasta kuehniella (Zeller)
Anomis (Cosmophila) flava (Fabricius)
Autographa californica (Speyer) Cryptophlebia leucotreta (Meyrick)
Cryptophlebia peltastica (Meyrick)
Diparopsis watersi (Rothschild) Diparopsis castanea (Hampson)
Earias insulana (Boisduval)
Earias biplaga (Walker)
Helicoverpa armigera (Hübner)
Helicoverpa zea (Boddie) Heliothis virescens (Fabricius)
Mamestra brassicae L.
Pectinophora gossypiella (Saunders) Spodoptera littoralis (Boisduval)
Spodoptera exigua (Hübner)
Spodoptera exempta (Walker)
Spodoptera frugiperda (Smith) Spodoptera su nia (Guénée)
- ' Syllepte derogata (Fabricius)
Hymenoptera Aphelinus albidopus (Hayat & Kausari)
Brachymeria olethria (Waterson) Chelonus curvimaculatus (Cameron)
Encarsia lutea (Masi)
Eretmocerus mundus (Mundus)
Gonozius sp.
Microbracon kirpatricki (Wilkinson)
Syrphophagus africanus (Gahan)
Spodophagus lepidopterae (Delvare & Rasplus)
Trichogramma brasiliensis Trichogramma lutea (G irau It) Fteteroptera Rhinocoris albopilosus (Signoret) Mites Polyphagotarsonemus latus (Banks)
Bacteria Bacillus thuringiensis
Agriculture et développement ■ SpiClCLÊIssitó - Novem ber 1997
harvest debris. This is done mechani-ca lly w ith a rotary cu ltiv a to r or by feeding the green plant parts to live-stock. Diapausing insects burrowed in the top soil layers can be efficient-ly destroyed by ploughing. Any chry-salides on the soil surface are gene-r a l l y d e s tgene-r o y e d by th e heat ogene-r predators.
In Africa, cotton plants are also often cut up and burned. It is essential to d e s tr o y all stems and r e g r o w th . Indeed, some pests (homopterans) can propagate on regrowth, w h ic h can also shelter in fe c tio u s agents (v iru s e s and p h y to p la s m s ) . R a t o o n in g , w h i c h is s o m e tim e s carried out by growers w h o have a shortage of seed, can enhance the possibility of pest propagation and infestations.
Pests can also be d is s e m in a te d through seeds. It is important to strict-ly c o n tr o l seed b eing tran spo rted from one ecological zone to another — especially to avoid disseminating pink bollworms.
Varietal characters
A plant's resistance to pests can be of morphological, biochemical or gene-t i c a l l y e n g in e e re d (i.e. a v a r ie gene-t y modified by introducing foreign plant genes).
M orphological characters
Various m o rp h o lo g ic a l plant traits function as physical barriers to pests or alter th e ir d e v e lo p m e n t c o n d i -tions.
Variations in leaf hairiness
Leaf hairiness is the most commonly promoted m orphological character. H a iry leaves ham per the d e v e lo p -m e n t o f jassid s, i.e. A f r ic a n
(J a c o b ia s c a spp.) and A sian
(.Amrasca spp.) species. The
efficien-cy o f th is t r a i t d e p e n d s on h a ir implantation patterns, length, shape and especially density (PARNELL et
al., 1949). Leaf hairiness is detrimen-tal to insects' feeding and oviposition behaviours. This can stall pest
infes-co to
p e s t
tvmaqmafc
tations, w hich is a major advantage during the vegetative stage of the cot-ton plant. Cotcot-ton therefore does not have to be sprayed w ith pesticides, and b e n e fic ia l insects (w h ic h are highly active during this period) will consequently not be endangered. Conversely, it has been reported that glabrous leaves can hinder oviposi- tions of some lepidopterans, espe-c i a l l y th e H e l i o t h i s v ir e s espe-c e n s /
H e t e r o p t e r a a r m ig e r a c o m p le x (BHAT et al., 1986). This character is
Pectinophoro gossypiella also useful for avoiding whitefly
out-on a cottout-on flower. breaks (GERLING, 1990), but can
Photo CIRAD-UREA
favour aphid development (DEGUI- NE, 1995). A c c o rd in g to W ILS O N (1986), under field cropping c o n d i-tions, this glabrous trait helps avoid cotton fibre contamination by plant debris, whereas the e nto m o log ica l advantages are questionable.
Lamina
The thickness, hardness and shape of the lamina — with "okra-type" laci- niate leaves — could have a role in pest resistance. Some varieties with laciniate leaves are now being crop-ped on large cotton plantations, e.g. cv Siokra in Australia and cv Sudae K in Sudan. These "okra-type" leaves e n a b le high a ir c i r c u l a t i o n , thus drying out pest larvae (e.g. B. tabaci) and enhancing pesticide penetration. D E G U IN E (1 995) fo c u s e d on the im p a c t o f th is c h a r a c te r on /4. gossypii outbreaks but found that it is not very e ffic ie n t. M o re o v e r, weed g ro w th is favoured w ith this type of leaf cover as a considerable amount of light reaches the ground.
Bracts
Atrophied or absent bracts, or those that are separated from the cotton boll ("frego" bracts), can hinter ovi- p osition ing by some lepidopterans and A. grandis infestations (ANGELI- Nl et a i, 1965; JENKINS, 1989).
N e ctary glands
As often pointed out, an absence of nectary glands can be beneficial, i.e. the cotton plant is not an interesting food source or attractive to certain insects (especially su cking pests). However, this character can be detri-mental since beneficial insects are also affected.
Conclusion
In cotton, p la nt breeders have not promoted many morphological plant characters that are known to hamper pest attacks because they are often negatively correlated with technolo-gical aspects of the fiber (PAULY & VAISSAYRE, 1980).
Biochemical characters
Various b io c h e m ic a l pest-resistant characters o f c o tto n plants — pH
D e l t a p i n e c u ltiv a r w i t h " f r e g ó " b r a c ts .
Photo CIRAD-UREA
Glossary
A n t i b i o s i s : the plant adversely affects the metabolism of the insect pest, sometimes to the extent o f killing it.
A n t i x e n o s i s : m o rp h o lo g ic a l or p h y s io lo g ic a l characters o f the plant that have a
repellent effect on insect pests.
A t t r a c t a n d k i l l : a concept that involves attracting the insect w ith a sex pheromone and then killing it w ith a contact pesticide.
C h e m i c a l m e d i a t o r : a v o la tile substance tha t p ro v id e s the insect w ith c ertain information, thus altering its behaviour w ith respect to host plant choices, egg-laying sites and sexual partners.
E n t o m o p a t h o g e n i c o r g a n i s m : a m icroorganism (virus, bacterium , fungus) that can cause diseases in insect pests.
H y p e r p a r a s i t e : a parasite that can be parasitic to another parasite.
G V : granulosis virus, where a single virion is incorporated in a protein body.
M a t i n g d i s r u p t i o n : spraying a specific pheromone upsets male detection of females, thus reducing mating.
NPV: nuclear polyhedrosis virus, where several viruses are incorporated in an protein body.
P a r a s i t o i d : an organism that lives through part o f its developm ent cycle w ith in the body of a host insect and subsequently causes its death.
P h y t o p l a s m : m ic ro o rg a n is m s c lo s e to b a c te ria th a t are p a th o g e n ic to plan ts (previously called mycoplasmal organisms).
P i b : polyhedral inclusion bodies containing active viral elements or virions that are released in the digestive tract of the host insect.
T o l e r a n c e : the plant is able to withstand pest infestations, w ith o u t being seriously damaged.
A griculture et développement ■ Sp^ClQÜ/Issitó - Novem ber 1 997
level of the cell contents, foliar tu r-gescence, the plant's sugar, protein and m in e r a l salt c o m p o s i t i o n — modify pest behaviour and develop-ment. Chemical substances (gossy- pol, ta n n in s , fla v o n o id s ) have an antibiotic effect on some pests such as H. a r m ig e r a , jas s id s and fle a beetles. In USA, many plant breeding studies are under way on these bio-ch e m ic a l traits, but no s ig n ific a n t results have been obtained to date (JENKINS, 1994).
In French-speaking Africa, there has been so m e success in b r e e d in g glandless cotton varieties, w ith the aim o f p ro m o tin g cottonseed and derivatives. M ore than 350 000 ha were cropped w ith glandless cotton in 1994-95. The problem is that the lack of gossypol glands weakens the p la n t 's n a tu ra l d e fe nse s, w h i c h means that pest m anagem ent staff have to be highly vigilant. Glandless c o tto n pla nts can be a tta cked by m an y d iffe r e n t insect pests at the onset of the growth cycle, especially coleopterans such as Halticinae spe-cies in A fric a and C h ry s o m e lid a e s p e c ies in S o u th e a s t A sia (BRADER, 1 9 6 7 ; G E N A Y , 1 99 4). M oreover, non-insect pests, w h ic h are rare in t r a d i t i o n a l c o tto n c r o p p i n g system s, can be present with glandless varieties: birds d u r in g p la n t in g and rod en ts and other m am m als d urin g the harvest period.
Transgenic cotton
Genes encoding toxins derived from the bacterium Bacillus thuringiensis can now be inserted in cotton plants th ro ug h genetic e ng ineering te c h -niques. Genetically modified varie-ties are now available on the market (PANNETIER et a i, 1995).
Toxins corresponding to the genes
Cry lA(b) and Cry IA(c) are active against the H. virescens/H.armigera c o m p le x and P. g o s s y p ie l la ( M c lN T O S H e t a l . , 1990). Recent tests by C IR A D h ig h l i g h t e d th a t
Cry IB has a similar spectrum of
acti-vity. Cry IIIA, a coleopteran-specific gene, could provide glandless cotton
varieties with resistance to coleopte- rans in the C h ry s o m e lid a e fa m ily (SEKAR et a i, 1987).
Problem s in v o lv e d in in tro d u c in g transgenic plants should not be over-shadowed by the positive aspects of these innovations (e.g. improvement o f c o m m o n v a rie tie s ). T he re is a considerable risk that some insects w il l q u ic k ly d ev e lo p resistance to
B. t h u r i n g i e n s i s to x in s (Me G A U G H E Y , 1 9 8 5 ; T A B A S H N IK , 1994). Research programmes are pri-marily focused on insertion of genes c o d in g for protease in h ib ito rs , or other factors (e.g. oxidases), w h ile attempting to m inimize any possible development of resistance in the tar-get pest. Secondly, p re c a u tio n a ry measures that should be taken when cropping transgenic plants are being investigated in terms of mosaic crop-ping patterns, crop alternation, and refuge c ro p p in g w ith a m ix tu re of transgenic and u n m o d ifie d plants (GOULD, 1995).
Entomologists are also closely inves-tigating how these genetic manipula-tions could upset the balance of the pest spectrum.
Entomophagous
insects
Entomophagous insect populations evolve th ro ug ho u t the agrosystem, i n c lu d in g c r o p fie ld s , fa llo w s and u n c u lt iv a te d host p lants g ro w in g near farms.
There have been many studies aimed at identifying and enumerating auxi-liary organisms in cotton cropping systems, while also investigating their roles in reducing pest p op ulatio ns and assessing unwanted side-effects of pesticides on these beneficial spe-cies. Indeed, v e ry little is k n o w n about the w id e variety of e n to m o -phagous insects.
Entomophagous insects have been inventoried in several African coun-trie s ( B u r k in a , C a m e r o o n , C ôte
Egg laying of
Trichogramma minutum on
Helicoverpa armigera egg.
Photo CIRAD-UREA
Spodophagus lepidopterae
preying on a Spodoptera littoralis larva. Photo CIRAD-UREA Apanteles sagax larvae on Syllepte derogata. Photo CIRAD-UREA
Cotton pest complexes
Cotton pest com plexes have been reviewed by several research teams, in clud in g: SILVIE et al., 1989; BOURNIER, 1991; DECUINE, 1991; LECOEUR & VAISSAYRE, 1991 ; B AG AYO KO et al., 1993; EKUKOLE, 1993; GALVA, 1993; SILVIE et al., 1993; STREITO, 1994.
A u x i l i a r y o r g a n i s m s o b s e r v e d o n H. armigera
A long list of auxiliary organisms have been identified for H. armigera but, as is the case for most pests, this list is not yet exhaustive. T. lutea, an oophagous pest o f the Trichogrammatidae fam ily, was recently identified in Burkina.
P a r a s i t o i d s o f S. littoralis
No oophagous parasites have been reported for 5. littoralis. In 1988, there was a m ajor discovery o f a new parasite, S. lepid op te rae (Pterom alidae), w ith unusual biological characteristics, w hich have been the focus of some attention (BOURNIER & BENMOUSSA, 1993; RASPLUS & DELVARE, 1994). Possible hosts were identified by CIRAD research teams, i.e. S. t'rugiperda, S. exigua, S. sunia and H. armigera.
P a r a s i t o i d s o f P. gossypiella a n d 5 . derogata
In c o n tin e n ta l A fric a and Madagascar, o n ly a fe w parasitoid species have been reported on P. gossypiella: Gonozius sp. (Bethylidae); Apanteles sp. (Braconidae);
B. o le t h r ia ( C h a lc id id a e ), M . k ir k p a t r ic k i (Braconidae), and C. c u rv im a c u la tu s (B raco nid ae ). A pest c o m p le x w ith a w id e range o f species was d e s c rib e d fo r S. derogata in Chad and Togo (SILVIE, 1991 & 1993).
P a r a s i t o i d s o f A. gossypii
M any different insects feed on A. gossypii. Concerning predators, DUVERGER (pers. com m .) drew up a map of Exochomus spp. distributions in Africa. DEGUINE (1995) pointed out the high relative numbers of ladybirds (adults and larvae) and especially
Cheilomenes spp. present in Cameroon. A. albipodus accounts for 56% of the overall p a r a s it o id p o p u la t io n in th e sam e c o u n t r y , w i t h S. a f r ic a n u s ( E n c y rtid a e ) a n d E n c a rs ia sp. ( A p h e lin id a e ) r e p r e s e n tin g 2 9 % an d 1 3 % , r e s p e c t iv e ly , of this population.
A u x i l i a r y o r g a n i s m s p r e y i n g o n B. tabaci
E x o c h o m u s sp. w as o b s e rv e d f e e d in g on B. t a b a c i w h it e f l y la rv a e in M a li (BAG AYO KO, 1989).
Green lacewing larvae prey on some w h itefly larvae, but very few entom ological field observations have been reported. A few parasitoids have no w been identified, all b e lo n g in g to the A p h e lin id a e fa m ily : Encarsia sp.; E. lu tea (M a li, C a m e ro o n );
E. transvena (Benin, Burkina, M ali); E. mundus (Burkina, Cameroon, Mali).
d'Ivoire, Mali, Chad and Togo), and in Paraguay (MICHEL & PRUDENT, 1987). There are reference c o lle c -tions at CIRAD (Montpellier, France), in some countries (e.g. Togo), and at the International Institute of Tropical Agriculture (I IT A) in Benin.
Data obta ine d on e ntom ophagous insects in French-speaking Africa are summarized in Table 2.
Effects of some secondary
host plants on their
population dynamics
Secondary host plants for bugs and A. gossypii were identified along the edges of co ttonfields (P O U TO U LI,
1 99 4; D E G U IN E , 1 995), but very little is known about their roles with respect to beneficial insects.
As early as 1974, PEYRELONGUE & BOURNIER reported four parasi-toids of larvae and one of E. insulana
n y m p h s on a m a lv a c e o u s p la n t
{A. a s ia t i c u m L.) in M adagascar. In B u r k in a , p a r a s ito id s fr o m
H. arm ig era on tomato are m ainly
Tachninidae species, whereas hyme- nopterans are generally noted on cot-ton. Low parasitism levels occur on tomato and cotton, i.e. less than 5% ( N I B O U C H E , 1 9 9 4 ). In T o g o ,
A p a n te le s spp. w ere observed on
Urena lobata (Malvacae) leaves rol-led by S. derogata caterpillars prior to p la n t in g th e c o tto n c r o p . In Cameroon, parasitoids of A. gossypii, usually found on cotton plants, were detected with various bug species on cultivated plants (okra, sorghum) and
Table 2 . Beneficial entomophagous insects identified in tropical Africa.
Country Num ber of genera or species identified Num ber of species identified
predators parasitoids hyperparasites parasites oophagous parasites of predators
Burkina 56 69 10 8 16
Cameroon 41 27 3 -
-M ali 16 12 - -
-Chad 34 63 14 1 5
Togo 28 45 10 1 12
Larval ladybird (Cheilomenes vicinia) preying on aphids.
PhotoJ.-P. Degu ine
w i l d p la n ts ( C a lo t r o p is p r o c e r a Ait.R.Br.).
In the dry season, substantial infesta-tions of parasites such as A. albipo-
dus have also been noted on other host plants. It is therefore essential to take bug species other than those that damage cotton crops into considera-tion when assessing the populaconsidera-tion dynamics of beneficial species. Prey that attract predatory bugs are also found on w ild plants g ro w in g near c o tto n pla nts (SILVIE e t a!., 1993; POUTOULI, 1994).
Role of auxiliary
organisms under natural
conditions
Auxiliary organisms were investiga-ted in untreainvestiga-ted fields.
Impact on some pest insect
populations
P O U T O U L I (1994) reported highly original data on oophagous toids of bugs in Togo, where parasi-tism by various species ranged from 13% to 76%. In Burkina, samples of
H. armigera populations were found
to have a lo w rate o f p a r a s itis m (1.4%), whereas pathogens induced high mortality (48.3%). In Chad and Togo, parasitism rates on 5. derogata for various years ranged from 1 8% to
4 7% on w o r m s and 2 1 - 2 3 % on p up a , w it h hig h h y p e rp a r a s itis m (70%).
S tud ie s w e r e u n d e r ta k e n in the Central African Republic, Cameroon and Chad on entomophagous insects that prey on /4. gossypii (VAISSAYRE, 1970; DEGUINE, 1995). On untrea-ted fields, the predator c o m p le x is dominated by ladybirds (45-85% of the sampled population), followed by syrphids (14-37%) and green lace- w in g s (up to 2 5 % ). Surveys in T o g o and B e n in r e v e a le d hig h lo c a l c o n c e n t r a t io n s o f som e families such as Hem erobiidae and Chamaemyidae. These data are com -parable to those obtained by MICHEL (1992 a & b, 1993) in Paraguay. The dynamics of auxiliary organisms on /\. gossypii in western and central A f r i c a are w e l l d o c u m e n t e d . Predators are active at the beginning of the crop cycle, then a fungal disea-se caudisea-sed by N e o z y g ite s fre s e n ii (E n to m o p h th o ra le s ) appears from August on. Parasitoids occur at the end of the crop season (September, October), but their levels seem to be too low to have a serious impact on pest populations at this phase of the crop cycle.
The percentages of parasitism noted in M a li on B. ta b a c i show ed that auxiliary organisms actually have a
Aphis gossypii first and second instars.
PhotoJ.-P. D eguine.
Spodoptera littoralis on a cotton flower.
Photo CIRAD-UREA
c o n s id e r a b le e ffe c t on c r o p pest populations: 9-23% for E. lutea and 6-25% for E. mundus.
The effects of predators have not yet been seriously q u a n tifie d , as their a c tiv it ie s are d i f f i c u l t to m o n it o r under natural conditions. The effects of this type of auxiliary organism are also l i m i t e d by p a r a s ito id s . Egg parasitism is h in de re d by specific behavioural activities, e.g. the male p r e d a t o r i a l bug R. a lb o p i lo s u s (R eduviidae) guards its eggs u ntil they hatch in order to ward off threa-tening parasitoids.
In addition, "opportunistic" predators (assassin bugs, spiders) sometimes attack other predators.
Entom o ph ag ou s insects g e n e ra lly seem to have a m o d e r a te e ffe c t, except in terms of egg parasitism (a mechanism that has not yet been ful-ly clarified). Hyperparasitic insects b ec o m e ac tiv e in the presence of
Table 3. Effects of various pesticide active ingredients on auxiliary organisms. (Source: SIGRIST eta!., 1994)
Active ingredient tested Dose (g/ha) Effects on ladybirds Effects on syrphids Effects on spiders alphacyperm ethrin 18 highly toxic highly toxic non toxic bifenthrin 30 highly toxic highly toxic highly toxic cyperm ethrin 36 highly toxic highly toxic highly toxic esfenvalerate 22 non toxic non toxic non toxic fenvalerate 60 moderately toxic non toxic highly toxic endosulfan 750 slightly to xic highly toxic highly toxic chlorpyriphos-E 4 5 0 non toxic highly toxic highly toxic chlorp yrip h o s-M 500 non toxic highly toxic highly toxic dimethoate 400 slightly toxic highly toxic non toxic isoxathion 350 slightly toxic highly toxic moderately toxic isazophos 200 highly toxic highly toxic non toxic om ethoate 300 slightly toxic highly toxic moderately toxic m etham id ophos 300 slightly toxic slightly toxic non toxic m o nocro lopho s 250 highly toxic highly toxic highly toxic profenofos 150 highly toxic moderately toxic highly toxic triazophos 125 slightly to xic highly toxic non to xic benfuracarb 250 slightly to xic slightly toxic moderately toxic carbosulfan 300 highly toxic highly toxic slightly toxic thidicarbe 800 slightly toxic slightly toxic slightly toxic im ida clo prid 50 slightly toxic slightly toxic non to xic
high primary parasitoid populations (e.g. S. derogata).
All research conducted on the effects of auxiliary organisms and their prac-tical interest should take these inter-actions into account.
Release of auxiliary
organisms
In Madagascar, tests began as early as 1971 on in t r o d u c in g a u x ilia r y organisms and rearing them for mass propagation and subsequent release in cotton cropfields (as part of a bio-logical control programme). The aim was to delay the tim e of the initial pesticide treatment, as some cases of pesticide resistance had been repor-ted (BOURNIER & PEYRELONGUE,
1973). The introduced parasite was a
T. brasiliensis strain from El Salvador and H. armigera was the target pest. The imported strain was propagated on a s u b s titu te host (the p y r a lid A. k u e h n ie lla ) before release. The results were not very encouraging as it was difficult to apply the technique in fa rm e rs ' fie ld s . S im ila r e x p e r i-ments were carried out more recently in Senegal (1979-1980), Togo and Cameroon (1982-1983) (SOGNIGBE, 1 9 8 9 ; B O U R N IE R , 1 9 9 1) u s in g insects from C IR A D 's M o n tp e llie r (France) laboratories.
A lth o u g h co ns ide ra ble skills have been developed for studying a u x i-liary organisms, it is still difficult to a p p ly the results on a large scale under tropical African cotton c ro p -ping conditions.
Impact of active
ingredients on auxiliary
organisms
Since 1990, the effects of pesticide active ingredients have mainly been investigated in Chad, Cameroon and Côte d'Ivoire. Despite the methodo-logical problems involved in these studies, an active ingredient classifi-c a tio n has been d ra w n up on the basis of their results (Table 3).
osÈÈmpssb w m ü û m M
Entomopathogenic
agents
Insect populations are controlled by various epizootic diseases involving viruses, bacteria, fu ngi and p ro to -zoans. Research undertaken in Africa on this to p ic was first reviewed by ATGER (1970). Bibliographical sum-maries have been published on lepi-do pte r a n v iru s e s and on u s in g
B. th urin g ie n s is as a biopesticide. M o r e r e c e n t resu lts o b t a in e d in Cameroon and Togo have also been pub lish ed (AN G ELIN I & JACQUE- MARD, 1 984; JACQUEMARD, 1987; M O N T A L D O , 1991 ; SILVIE et al., 1993).
Insect viruses
W h e n c o n d it io n s are s u ita b le for their development, epizootic viruses can have spectacularly destructive effects on some lepidopteran larval pop u la tio n s , e.g. S. exem pta o u t-breaks in grass crops (Africa) and
A. argillacea in cotton crops (Latin
America).
Virus infections have been detected in most lepidopterans found on cot-ton crops in Africa, i.e. in major pests such as H. a rm ig e ra , D. w a te rs i,
E. insulana and C. leucotreta, and in secondary species such as S. exigua,
Table 4. Potential use of commercialized insect virus diseases (RIBA & SILVY, 1993). Pathogen Brand name Origin Potential target (species)
Heliothis NPV Elcar Viron H Biotrol VZH Switzerland USA H. armigera H. armigera USA H. armigera Spodoptera NPV Spodopterin Viron P Biotrol VPO France USA USA S. littoral is S. littoral is S. littoral is Mamestra NPV Mamestrin France H. armigera Autographa NPV MGS 400 USA H. armigera Cydia GV SAN 406 Carpovirusin Granupom Decyde Switzerland France Germany USA C. leucotreta C. leucotreta C. leucotreta C. leucotreta A m s a c ta sp. (A N G E L IN I & V A N - D A M M E , 1 9 6 9 ; C RO IZIER et al., 1983; ANGELINI & JACQUEMARD, 1984).
V irus infe c tio n s w ere observed in
A. flava in Mali and in S. littoral is in Chad (ATGER & CHEVALET, 1975; ATGER, 1970). H ow e ver, no virus infec tio ns have been noted in the b o llw o r m P. gossypiella or in the phyllophagous caterpillar S. derogata. There have been several attempts to propagate locally isolated viruses (in Chad and Côte d'Ivoire), and finally e x p e r im e n ts w e r e c o n d u c te d on using viruses isola ted fro m o th e r insects — at lower cost than w ou ld have been possible with locally pro-duced viruses. The most com m on ly used p a th o g e n s are n u c le a r p o ly h e d r o s is viru s e s (NPV) fro m
/ \ . C a l i f o r n i a and M . b r a s s i c a e
(JACQUEMARD & DELATTRE, 1977; JACQUEMARD, 1978).
Some e n t o m o p a th o g e n ic viruses have been mass-reared in vivo from insect tissue cultures and marketed by pesticide manufacturers. CIRAD studies revealed that these products can be used to control some cotton pests (Table 4).
Research scientists have tried to utili-ze possible synergetic interactions between e ntom opathogenic agents and c h e m ic a l p e s tic id e s — often pyrethroids at low doses (FERRON et
al., 1 983). Studies on this feature were conducted in Cameroon, Côte d'Ivoire, Chad and Togo (JACQUEMARD, 1982; RENOU, 1987; M O N -TALDO, 1991; VAISSAYRE, 1994).
Treatment of cotton plants
B io p e s tic id e trea tm en ts o f c o tto n plants w ith these insect viruses are efficient when the control operator complies w ith a treatment schedule and a dose of 1 0 13 Pib (polyhedral in c lu s io n bodies) per hectare fo r m ost o f the b a c u lo v iru s e s tested (CAUQUIL, 1985).
However, this efficacy is limited by: - the biopesticide persistence, which is often insufficient because of the presence of antagonistic factors such as UV-rays or fo lia r secretions and their pH levels;
Vi rus-infected caterpillar on a cotton leaf.
Photo CIRAD-UREA
- the fact that the agent generally has to be ingested in order to becom e active, which is always difficult with bollworms (but phagostimulants can be added to enhance ingestion); - t h e specificity, i.e. requiring propa-gation of the strains so as to control the constantly changing nature of the parasitism;
- the problems that arise in obtaining suitable quantities of the virus, which is propagated on reared insects or tissue cultures.
The most recent "combined control" strategy involved associating a poly-h e d ra l v iru s ( u s u a lly N P V fr o m
M. brassicae ), a phagostimulant and a lo w dose of c h e m ic a l pesticide. The results in controlling certain boll-worms (H. armigera and D. watersi ) were as good as can be obtained with chemical pesticides (RENOU et a i, 1 9 8 5 ; SILVIE e t a l. , 1 9 9 3 ). Nevertheless, possible incompatible interactions (depending on the pesti-c id e used) have to be taken in to account. C o m b in a tio n s w ith p yre-throids generally give suitable results, w he re as the use o f some o rg an o- phosphorus compounds (monocroto- phos) can lead to a loss of efficacy. In a d d itio n , the high cost of treat-ments, such as those recently tested in Cameroon and Togo, is a conside-rable c o n s tra in t fo r the extension o f m i c r o b i o l o g i c a l te c h n iq u e s to control bollworms.
Entomopathogenic
bacteria
In the e n to m o p a th o g e n ic bacteria group, there are many opportunistic agents that can multiply to the extent of killing the host. Injuries generally provide a port of entry for these bac-te ria . H o w e v e r , o n ly s p o r u la tin g bacilli can infect healthy insects.
Infection mechanisms
of identified bacteria
Most studies undertaken to date on e n to m o p a th o g e n ic b a c te ria have focused on B. thuringiensis (ARON-SON et a i, 1986; NBIAP, 1995). This g r a m + b a c t e r iu m fo rm s c ry s ta ls d urin g sporulatio n, w h ic h contain
toxins such as the 9-endotoxin, i.e. the most important and only legally- applicable species. It is released after ingestion by the insect and binds to membrane receptors of the midgut, which is subsequently destroyed. Initial investigations highlighted the activity of the Anduze strain against
E. ins u lan a (LE GALL, 1957; BUR- GERJON & GRISON, 1959). Other strains of this baccillus were then iso-lated from D. watersi, E. insulana and A. m o lo n e y i (JAC QU EM AR D , 1 9 6 5 ; ATG ER & J A C Q U E M A R D , 1965). Variations in the virulence of d iffe r e n t strains o f this b a c te riu m against Earias sp. were noted in labo-ratory tests (FRUTOS e t al.t 1987). All tested treatments conducted in co t-ton fields with B. thuringiensis form u-lations revealed the high efficacy of this bacterium against some pests, whereas it was found to be a poor biopesticide against bollworms (JAC-QUEMARD, 1987).
Other bacterial strains have been iso-lated (S e rra tia , P s e u d o m o n a s ,
A e r o b a c te r) from diseased insects
(Diparopsis, Heliothis, Spodoptera),
but their biopesticide efficacy has not yet been tested. Hence, B. t h u r in
-giensis is the only bacterium that has been used as a b i o p e s t i c i d e to control leaf-eating caterpillars, espe-cially S. derogata and A. flava.
Prospects
Around 10 toxins from B.
thuringien-sis s tra in s have c u r r e n t l y been c h a ra c te riz e d . Genes that e n a ble their synthesis have been identified, with their specificities determined in le p id o p te r a n s and c o le o p te ra n s . A collection of toxins with high effi-cacy against cotton pest insects is available, and includes: Cry IA(b),
Cry IA(c), Cry IB and Cry IIA genes
for bollworms, Cry 1C for Spodoptera sp. and Cry III for Chrysomelidae, pests of glandless cotton.
A cotton im p ro v e m e nt programme has been set up to develop transgenic cotton plants expressing B. th u rin
-giensis toxins. This marks an impor-ta n t tu r n in g p o in t in the quest to co n tro l co tto n b o llw o rm s that are relatively unsusceptible to standard
B. thuringiensis spore and protein crystal
Photo CIRAD-IG EPAM
o s ttm p& sb
biopesticide treatments.
The role of entomopathogenic bacte-ria in integrated pest m anagem ent p ro g ra m m e s s h o u ld th e r e fo r e be reconsidered. B. thuringiensis formu-lations could be used as pesticides against phyllophagous caterpillars by d e t e r m i n i n g th e m o s t s u it a b le spraying techniques, possible asso-ciations with chemical pesticides and the best times, relative to the insects' d e v e lo p m e n t c y c le , to c o n d u c t treatments (e.g. it is best to spray just- hatched neonate caterpillars) (DABI, 1988; HUSSEIN et a i, 1990; PLAPP, 1991). Transgenic varieties w ill have to be d e v e lo p e d to c o n t r o l b o l l -w o r m s (B E N E D IC T e t a i , 1 9 9 2 , 1993; GATEHOUSE et al., 1 992).
Entomopathogenic fungi
E p iz o o t ic diseases ca u s e d by En to m o ph th orale s can be respon-sible for killing off insect populations under natural conditions. This phe-nomenon was noted for aphids in tro-p ic a l A fric a (SILVIE &amtro-p; D E G U IN E , 1994). H ow e v e r, spray treatm ents w ith com m ercial mixtures, such as M y c a r ( H irs u te lla th o m p s o n ii) or Vertalec (Verticillium lecanii) against
P. latus, aphids and whiteflies, res-pectively, have failed. This was likely due to the fact that the most suitable ecological conditions for the action
of these pathogens are still not clear-ly established.
Chemical mediators
For several years, sex attractants have been used to tra p l e p i d o p t e r a n insects. CIRAD, in collaboration with the Chemical Mediator Laboratory of the French In s titu t n a tio n a l de la recherche agronomique (INRA), have developed a num ber of sex attrac-ta nts such as p h e r o m o n e s fr o m
H. armigera, C. leucotreta, E. insula-na and D. watersi (DESCOINS & GALLOIS, 1979; A N G E LIN I et a i , 1976, 1980, 1981). On the basis of this work, synthetic sex pheromones were used to investigate the popula-tio n dyna m ics o f a du lt males, e.g.
H. armigera, C. leucotreta, S. littora-lis and P. g o s s y p ie l la (DAIBER, 1 9 7 8 ; B O U R D O U X H E , 1 9 8 2 ). U nfortun ately, the results of many tests carried o ut in seven d ifferent countries are not yet very useful for a g r ic u ltu r a l w a r n in g purposes or d e c is io n - m a k in g on pest c o n tr o l treatments (JACTEL & VAISSAYRE, 1988; MICHEL, 1992a and b).
In tropical African peasant farming conditions, mating disruption te ch -niques are difficult to apply because of the fragmentation of cotton fields, staggered planting and fruiting dates, and technical treatment constraints.
Trapping and mating
disruption
Studies conducted by CIRAD on this topic have focused on several diffe-rent species.
H. armigero
The population dynamics of H.
armi-gera were investigated, but very few or no correlations between captured males and crop damage in the field w e re n o te d . Eggs w e re o b s e rv e d before the adult males were captu-red, suggesting that the cotton fields are a c tu a lly c o lo n i z e d by g ra v id females. The usefulness of trapping males to determine when pesticide treatments should be undertaken is therefore questionable.
Map of receptor sites (located around intestinal microvilli) of the Cry IAa toxin (^ ) in the midgut of Chilo suppressolis.
Photo L. Fiuza
Diparopsis watersi.
Photo J.-P. Degum e Moreover, mating disruption w ould
not be a pplicable for this polypha- gous ubiquitous pest.
Tests on D. watersi
Very few tests have been carried out on D. watersi, apart from monitoring a du lt pop ulatio ns. It is possible to attract the first upsurging males of this m onophagous pest to an early planted or pruned cotton trap-crop. A mating disruption or attract and kill tr a p p in g s trategy c o u ld be used, c o m b in in g a sex p h e ro m o n e w ith glue or a pesticide.
E. insulana
and E. bip la g a
No studies have focused on E.
insula-na and E. biplaga, despite the fact
that these tw o species (w h ic h are often found on the same sites) cause 20-30% of all caterpillar damage to cotton fruiting organs. For these oli- gophagous species, the mating dis-ruption strategy w o u ld certainly be valid and readily possible. O n ly the
E. insulana sex pheromone is c o m -m ercially available. Moreover, this pest species can be quite easily rea-red on a r t i f i c i a l m e d iu m , w h i c h means that more advanced studies are would be possible.
P. gossypiella sex pheromone
The sex pheromone of P. gossypiella (g o s s y p lu re ) has been p r o d u c e d
industrially for about 15 years. The results of many studies on diffusion of this pheromone have led to a w id e v a rie ty o f s o lu tio n s ( p h e ro m o n e - soaked microfibres, tape or cord, and spraying water-suspensions of micro-g ra n u le s ), th a t can be c h ose n to address d if fe r e n t o b je c tiv e s , e.g. investigating population dynamics, mating disruption and, more recent-ly, attract and kill (F1ENNEBERRY et
a i , 1981; CRITCHLEY et al., 1983;
HOEER & BRAZZEL, 1992). Mating d is r u p t i o n is a p o p u l a r c o n t r o l method, but the results have been quite variable (USA, Egypt, Pakistan). Interesting test results were obtained at Bouaké (Côte d'Ivoire): damage to fruiting organs was m arkedly redu-ced by m a n u a lly s p ra y in g c o tto n leaves w ith glued microtubes (SAN- DOZ), at 5 000-8 000 microtubes/ha (VAISSAYRE, 1987) or using im pre-g n a te d tapes ( O C H O U , 1 9 9 7 ). Persistence was higher than obtained w ith m icrogranulated form ulations a p p li e d w i t h s ta n d a rd s p ra y in g equipment.
C. leucotreta
There is a p r o b le m o f s p e c ific ity w h e n c o n d u c tin g e x pe rim en ts on C. leucotreta, as males of other spe-cies are also captured in the traps, especially C. peltastica. There is no evidence that this lepidopteran could be controlled by mating disruption. This technique could be com bined w ith that used to co ntrol P.
gossy-pie lla in areas jo in tly colonized by these two endocarpal bollworms.
S. littoral is and S. derogata
For 5. littoralis, sex pheromones of various origins can be used to m oni-t o r a d u loni-t p o p u l a oni-t i o n d y n a m ic s . CIRAD has not been too involved in such studies because of the low eco-nomic impact of this pest in tropical Africa — more research is under way on this topic in Egypt.
A sex attractant was synthesized for
S. d ero g a ta , but it is not yet being
m a r k e te d ( H I M E N O & H O N D A , 1992). It is still quite difficult to rear this lepidopteran pest in an artificial e n v i r o n m e n t , as it seems to require a natural e n v iro n m e n t for ovipositioning.
Photo CIRAD-UREA
Bene/LCUJiSs
Pentatonid bug infected by the Beouveria fungus.
Photo CIRAD-UREA Coccinellidae eggs in an aphid colony. Photo CIRAD-UREA Phonoctonus, predator of Dysdercus. Photo CIRAD-UREA Coccinellidae larva feeding on an aphid colony. Photo CIRAD-UREA
Lacewing larvae feeding on whitefly pupea.
Photo CIRAD-UREA
Spodoptera larva killed
by Bacillus thuringiensis.
Future research on
chemical mediators
Pheromones: availability and
applications
Sex pherom ones are not yet avai-lable for three species (D. watersi,
E. biplaga and S. derogata ), and the-re athe-re no clear corthe-relations between trapped insects and field infestations, especially with respect to H.
armige-ra.
Mating disruption is only possible for o li g o p h a g o u s or m o n o p h a g o u s insects (P. gossypiella). Pests often c o lo n iz e the same fields, i.e. both exocarpal species (e.g. H. armigera,
Earias spp.) and endocarpal species (e.g. P. gossypiella, C. leucotreta ).
Field applications
It would be very difficult to set up a network of traps around cropfields, even if o rg a n iz e d by c o m m u n ity
associations (village comm unities). S im p le in e x p e n s iv e t e c h n iq u e s requiring little labour input should be developed, especially for app lyin g pheromones to induce m ating dis-ruption. The low number of emission sources and the persistence o f the formulation should be future reseach focuses.
Regardless of the results of previous operations, the success of such pro-jects w ill depend on how easily phe-romones can be applied in traditional farming systems.
Conclusion
The intensity and complexity of most pest infestations certainly warrants th e use o f c h e m ic a l p e s tic id e s . Treatments should, nevertheless, be c o nducted as part of an integrated pest m anagem ent p rogram m e that
also involves preventive measures, adapted cropping practices, rational choices of associated crops, and an overall understanding of the entomo- phagous insect c o m p le x (Table 5). Integrated pest m a n a g e m e n t p r o -grammes for cotton cropping systems should also inc lu d e destruction of harvest debris, tillage, intercropping, and well chosen planting dates and positions of the crop in the cropping plan.
It is essential to choose the most sui-table cultivar to be cropped, i.e. phy-siological characteristics (length of its growth cycle, hardiness, compensation potential) and some m orph olo -gical traits (especially leaf hairiness). Integrated pest m a n a g e m e n t p r o -grammes imply respect for the bene-ficial fauna. However, the results of mass releases o f e n to m o p h a g o u s auxiliary insects (e.g. trichogrammids for biological control) have generally
Table 5. Com bined methods to control the main cotton pests as part of an integrated pest management strategy.
Pest Cultural control Varietal choices Biological control (1) Pheromones Chemical control
H. armigera * planting date * transgenic
varieties (B. thuringiensis)
** NPV: * trapping ***
C. leucotreta * planting date * transgenic varieties ** NPV: * trapping *** P. gossypiella *** destruction of harvest debris * atrophied or frego bracts *** mating disruption
Leaf-eating insects ** B. thuringiensis ***
Jassids ***
hairy leaves Aphids * destruction of
plants after harvest, topping, defoliant ** natural effect of beneficiais and entomopathogens ** seed treatments W hiteflies * destruction of plants after harvest, topping defoliant * "okra"leave s **
M irids * nectari less variet ies
Bugs Mites
***
***
*: established method; **: technique used on a reduced scale or w ith limited efficacy; ***: w idely used technique of acknow led-ged efficacy.
(1): entomophagous insects and entomopathogens are included in the biological control category.
Sex pheromone-soaked cord attached to a cotton plant.
Photo P. Silvie
n o t been v e ry e n c o u r a g in g . It is important to have an overall unders-tanding of the population dynamics of the main entomophagous insects present in a region, so that they can be taken into c o n s id e ra tio n w he n m aking decisions on active ing re -dients to be used in pest control treat-m ent progratreat-m treat-m es. Sotreat-me e n to treat-m o - p a th o g e n ic o rg a n is m s e f f i c i e n t l y reduce pest pop ulatio ns and treat-ments are therefore unnecessary (e.g. Entomophthorales against aphids). In some cases, microbiological prepara-tions can be app lie d (e.g. viruses,
B. thuringiensis) to control lepido- p teran pests. T re a tm e n ts a g a in st b o l l w o r m s are m o re l ik e l y to be successful when conducted at care-f u l l y chosen tim e s. C o m b in in g a pathogen with a low pesticide dose, i.e. a "c o m b in e d c o n tro l" strategy, w il l often enhance the e ffic a c y of the pathogen.
Apart from gossyplure (used to dis-rupt P. gossypiella mating) and the aggregation pheromone of A.
gran-dis, sex pheromones are not directly useful for pest control — however, they do provide informative results that could help in making treatment decisions. It should not be assumed that future transgenic cotton varieties expressing B. thuringiensis toxins
w il l d e fin itiv e ly solve all pest and disease problems with respect to this crop. Nevertheless, w ith the in tro -duction of such new varieties, insect populations w ithin cropping systems could be managed differently, unless they develop resistance to the toxins involved.
In c o n c lu s io n , the results of these alternatives to chemical pest control are still not conclusive enough to be e x p lo ite d . There are s o lu tio n s for co ntrolling some types of pests, for instance: leaf hairiness against j as-si d s ; e n to m o p h a g o u s insects and entomopathogenic agents can redu-ce h o m o p te ra n and le p id o p te ra n pests; and some transgenic cotton varieties are efficient in c o ntrolling b o llw o r m s . H o w e v e r, su bstantial crop damage w ill still occur unless c h e m ic a l p esticide treatm ents are used. To be e c o n o m ic a l ly sustai-nable in the short term, cotton crop m a n a g e m e n t s h o u ld be based on alternative cultural, varietal and bio-logical methods, always c om bined w ith supervised ch em ic a l co n tro l. This is the current strategy adopted by CIRAD, through the development of new cotton pest management pro-grammes in c o lla b o ra tio n w ith its partners.
Damage to a young flower bud caused by Earias biplaga.
PhotoJ.-P. Bournier
