t .
New cotton protection
programmes: limitations
and prospects
From 85% to 90% of all cotton cropland in French-
speaking Africa south of the Sahara — i.e. more than
a million hectares — are being sprayed with chemicals.
New pest management programmes have been
developed that involve treatments with lower doses
of active ingredients, but they are still as effective as
systematic preventive treatments. Attempts at wide-scale
application of these programmes have unfortunately
been hampered by socioeconomic constraints.
Cotton research institutions are trying to facilitate
implementation of these programmes through active
involvement in training projects.
P
est management accounts for m o re th a n h a lf o f all in p u t costs fo r co tto n p ro d u c tio n . This expense has been increasing sin-ce the w ith d ra w a l of subsidization (on fertilizers and pesticides) in 1992 and the devaluation of the CFA franc in January 1994. In addition, the pro-ducer price for seed cotton has drop-ped and the pesticide distribution has been privatized. This has had some surprising effects on prices, e.g. in Benin (1993) where the per-hectare treatm ent cost was fo urfo ld higher th a n w i t h p r o d u c ts p u r c h a s e d abroad. J. CAUQUIL, M. VAISSAYRE CIRAD-CA, BP 5 0 3 5 , 3 4 0 3 2 M o n tp e llie r C e d e x I , FranceAlthough farmers are becoming more technically skilled, this price increase has prompted them to adopt extensi fication strategies, as shown by the e x c e s s iv e l o w e r i n g o f p e s tic id e d osages fo r c r o p tr e a tm e n ts . Pesticides are bought at a fixed per- hectare price and usually on credit. M oreover, to reduce expenses, far-mers often deliberately underestima-te the area they actually crop, and spray th eir fields w ith o v e rd ilu te d m ixtures or reduce the n um be r of treatm ents. Surveys c o n d u c te d in B u r k in a /C a m e ro o n , Côte d'Ivo ire , Mali, Togo, etc., highlighted that far-mers do not comply with
d a tio n s p ut fo rw a rd by e xten sion agencies, instead they adapt them to their own financial situations.
Training and technical consultancy on cotton cropping are increasingly provided through village community o r g a n iz a tio n s — this reduces the need for a high number of technical field staff whereas the role of supervi-sory agents from these organizations is essential.
Pest management
programmes
adapted to different
situations
In each country, this standard pro -gramme has been streamlined such that pesticide treatments are adapted to farmers' technical skills, the social environment, the crop-yield potential and pest and disease conditions. Staggered-targeted control has been recommended for the northern part o f the c o tto n aera, w h e re ann ua l rainfall is less than 1 000 mm and exocarpal bol I w orm s (Helicoverpa
a r m ig e r a , D ip a r o p s is w a te rs i) are v e ry c o m m o n (D E G U IN E et al., 1993). This program m e involves a series of six treatments (on average), conducted on a 14-day basis, w ith low doses of one or tw o active ingre-dients. Infestation levels of the main pests are checked 7 days after each treatment: bollworm s susceptible to pyrethroids; Syllepte derogata and mites s u s c e p tib le to o rg a n o p h o s -phorous acaricides; Aphis gossypii and Bemisia tabaci, sucking insects
Agriculture et développement ■ Sp&XCtü'Lssitó - Novem ber 1997 Data sheet for pest counts.
Photo CIRAD-UREA
C o s t-e ffe c tiv e and e n v ir o n m e n t - friendly pest control solutions have now been developed. However, far-mers have to be trained on using the-se new techniques. Standard cotton pest management involves four to six p re s c h e d u le d u l t r a - l o w v o lu m e (ULV) treatments (i.e. 1-3 l/ha of pre-pared p ro d u c t, d e p e n d in g on the country). The first treatment is carried o u t 4 5 -6 0 days after sow ing, w ith subsequent treatments every 14 days.
susceptible to organophosphorous aphicides.
It is difficult to assess outbreak levels in the region infested with the endo- phagous b o llw o rm s P e c tin o p h o ra
gossypiella and Cryptophlebia leu-
c o tre ta , and in te rm e d ia te c o n tro l
programmes are thus used. There is about 300 000-350 000 ha of cotton cropland in this zone (30-35% of the overall area controlled).
Binary mixtures of a pyrethroid and an o r g a n o p h o s p h o r o u s a c a r ic id e (at lo w doses), w h e n th e y have a potentiation effect, could be recom-mended in areas where Syllepte and b o llw o rm s are both present (C A U -QUIL & VAISSAYRE, 1994). For ins-tance, with cypermethrin and triazo- phos blends, the standard doses of 30 g/ha for cypermethrin and 150 g/ha for triazophos could be reduced to 15 + 75 g/ha, respectively. If pest populations are high, an additional treatm ent has to be added 7 or 14 days after the first scheduled treat-m e n t. E x te n s io n a g e n c ie s in Cameroon have opted for this latter s o lu tion w hereas the first one has been adopted in Benin, Burkina and Mali.
A mixed programme is being tested in Côte d'Ivoire, including threshold- based treatments at the beginning of the crop cycle, calendar scheduled
Counting crop pests with a pegboard in Burkina. Photos CIRAD-UREA
tr e a tm e n ts to c o n tr o l b o l l w o r m s (exophagous and endophagous) and o p t i o n a l a d d i t i o n a l tr e a tm e n ts , depending on observed levels of suc-king insects and mites.
Intermediate control programs are of c o n s id e r a b le in te re s t in te rm s o f reducing risks, as treatments are pres-cheduled. M oreover, this provides an efficient w ay of training farmers on h o w to assess pest in fe s ta tio n levels and dete rm ine w h a t type of product and dosage is most suitable under the prevailing conditions. W ith this tra in in g e xperience, fa r-mers could later switch to threshold- based treatments. Attempts at promo- tin g w id e s p re a d use o f th re s h o ld interventions in Notsé (Togo; 1989- 1991 ) and Garoua (Cameroon; 1992) revealed that a tran s ition phase is needed to give farmers time to master this new strategy.
Pests, crop damage
and treatment
thresholds
Farmers generally consider that all insects are pests, w ithout perceiving th a t some insects are b e n e fic ia l. SIGRIST (1992) found that Cameroo-nian farmers separate insects into two groups, i.e. those that crawl (cater-pillars, w h ite w orm s , etc.) and fly (butterflies and bugs). They act when c ro p d am age is v is u a lly o b v io u s : withering (e.g. Fusarium wilt), virus diseases (mosaic virus), rolled leaves due to 5. derogata, and sh rivelled leaves after attacks by sucking insects are all considered to be more serious than outbreaks of endocarpal b o ll-worms, for instance, w h ic h are not externally visible.
Insects are well perceived in terms of the damage they cause, e.g. aphids and honeydew, bollworm s and abs-cission of fruiting organs or boll per-foration. However, farmers should be trained on mites, mirids, vectors of biologically transmitted diseases.
Pest infestation levels are assessed from specific-sized samples
(general-ly 25 plants/unit plot) by counting the number of pests or noting the dama-ge they have in d u c e d . These esti-mates mainly apply to mites, sucking insects, leaf-eating caterpillars and bollworms.
As mites are very tiny, outbreaks are evaluated in terms of damage they cause. Curative treatments should be undertaken once the first leaf damage is observed and related with weather conditions, as mite populations can expand very quickly.
Aphid colonies are generally found on te r m in a l s h o o t leaves. A t the beginning of the growth cycle, infes-tation levels can be evaluated on the basis o f the n um ber of plants w ith shrivelled leaves — a sign of aphid attacks.
W h itefly (B. tabaci) populations are estimated by c o u n tin g adults very e a rly in the m o r n in g on te rm in a l shoot leaves or nymphal forms with a m a g n ify in g glass. These n y m p h a l counts are done on leaves chosen according to specific criteria, such as number, area, position, and growth stage.
The leaf roller S. derogata is the most com m only reported leafeating cater-p illa r . O u tb re a k s are assessed by c o u n tin g the n um ber of w orm s on each plant, w h ic h is preferred over counting rolled leaves (or plants with rolled leaves), as these symptoms are persistent.
Exophagous bollworms (H . armigera,
D. watersi, Earias sp.) are visible on the external parts of plant organs, and population growth can be evaluated by w o r m counts. In contrast, o u t -breaks of endophagous bollworm s, which occur on more than one-third of all cotton cropland, can only be assessed by random cracking of frui-ting bodies.
Cotton fields are surveyed w eekly, w it h o b s e rv a tio n s n ote d on data sheets (C a m e ro o n ) or p e g b o a rd s (Burkina, Benin, Chad). One major advantage of pegboards is that they can be used by illiterate scouts,
reas observations have to be written d o w n and summed up on the data sheets. Nevertheless, data sheets are of considerable interest as the quality of the m onitoring w ork can be eva-luated, data can be compared for a g iv e n z o n e to h e lp in d e c is io n m a k in g , and n o m o g ra p h s can be used w h e n the im p a c t o f the pest varies at different crop stages. The latter occurs with aphids, w hich are considered to be much less threate-ning in the middle of the crop cycle than at the beginning or the end — treatment thresholds therefore vary with the extent of risk.
Pest-specific treatment thresholds are determ ined on the basis of e c o n o -mic, climatic, biological and agricul-tural factors. They generally do not vary markedly between countries. It is cru c ia l to c o n d u c t treatments immediately after the threshold has been reached, and the choice of pes-tic id e and dose is also im p o rta n t. Special measures should be taken during serious pest outbreaks, espe-cially for H. armigera.
can easily monitor drum recycling as they are not supplied in large quanti-ties.
Toxicant monitoring
African farmers' conception of pesti-cides is h ig h ly c o m p le x : th ey are considered as poisons with a m ythi-cal a ura (a p r o d u c t fr o m the " m o d e r n " w o r ld ) w h i c h can also have other uses (for fishing, hunting, etc.). T raining is essential as many farmers are unable to decipher and interpret pesticide labels, or they do not abide by the instructions (which, in any case, are often not sufficiently detailed).
Pesticide packaging has to be well d es ig n e d to a v o id p o l l u t i o n p r o -blems. Each kind of packaging has advantages and drawbacks, depen-d in g on the u lt im a t e use. M e ta l drums are solid and enable storage of small volum es of product, but it is d iffic u lt to pour out the contents in small batches (tap, funnel, scoop). In addition, these drums are comm only reused, e.g. as distillation equipment, for transporting water, roofing
mate-rial and kitchen use. Extension agents
A griculture et développement ■ SpeCtCtü/Lssitó - Novem ber 1997
VLV spraying equipment. Photo CIRAD-UREA
Various types of pesticide packaging. Photo CIRAD-UREA
Plastic cans are readily stored, trans-ported and handled. Farmers mainly reuse them to hold edible liquids or fuels. The results of a survey carried out by SIGRIST (1991) of 73 villages in Cameroon indicated that 19% of these containers are destroyed after use, 18% are reused for fuels, and 63% for food products (water, oil, honey, cowpeas). In Côte d'Ivo ire , CHEYDA (1991 ) noted that all plastic cans w e re r e c y c le d after rin s in g , whereas metal containers were des-troyed or discarded.
M e ta l or h e a v y - d u ty s in g le -d o s e containers reduce handling and the risks o f m a k in g tr e a tm e n t dosage errors, but the per-litre purchase pri-ce is higher. As these containers are relatively small, they almost never reused.
In French-speaking Africa, "English packs" — including different form u-lations, instructions in the local lan-guage, gloves, and sometimes even a mask — are never used. The problem is that farmers fields are often larger than o f f ic ia lly declared and these packs are designed for treatment of a specific surface area.
C o n tr o l scouts have to be h ig h ly observant to ensure the efficacy of these n ew pest m a n a g e m e n t p r o -grammes. These scouts choose repre-sentative samples, and carry o ut a series of counts to assess populations of the main pests. The technician that does this w o r k should also aim at training head farmers or members of their families. In Cameroon, control scouts are often youths who have just fin ish ed th e ir s c h o o lin g . They are paid by the local village communities for the task of monitoring an area of about 40 ha (cropland concentrated in blocks). In western Africa (Benin, Burkina, etc.), cropland is relatively scattered and farmers do not wish to have their fields surveyed by external agents. It is therefore necessary to train a member (preferably literate) of the farmer's family.
The scout always has to be attentive and his job is repetitive and difficult on account of the morning dew and heat d u r in g the day. The tra in in g supervisor has to keep checking the trainees' w o rk , at least d u rin g the a p p re n tic e s h ip (in C am ero on , the head control staff regularly co lle c t copies of the scouts' data sheets). To warrant their positions, these supervi-sors have to have a high level of tech-nical and training skills. In addition, the "neighbouring extension" pheno-m en on can be used at o p p o r tu n e times, as fa rm ers are often h ig h ly receptive to innovations being tested nearby.
Training steps
Farmers often do not comply with the p r e c a u tio n a r y r e c o m m e n d a tio n s , especially concerning protective clo-thing. In Côte d'Ivoire, a survey by C H E Y D A (1991) revealed th at all control operators wore short-sleeved shirts or t-shirts, but 20% were wea-ring shorts; only 5% wore boots, 10% laced shoes, 20% sandals, and 65% were barefooted; finally, 90% were bareheaded. All questioned opera-tors stated th at th ey w ashed th e ir hands, but usually did not change th e ir c lo th in g after a treatm ent. In C a m e ro o n , SIGRIST (1991) noted that control staff did not eat, drink or smoke during treatments; they some-times protected their heads, took the w in d d ir e c t io n in to a c c o u n t and washed completely after control ope-rations.
Training aids
Training is not a top research objecti-ve, b ut it is essential for e ffic ie n t implementation of research results. For extension purposes, a conside-rable effort has to be made to train field workers on new crop protection programmes w hile highlighting their economic impacts.
The tre a tm e n t th re s h o ld c o n c e p t, based on th e a ssessm ent o f pest dam age fro m an e c o n o m ic v i e w -point, should be stressed at the outset when training farmers on new pest m a n a g e m e n t p ro g r a m m e s . This represents a new intellectual challen-ge for many farmers w h o might, for instance, be hesitant to begin a treat-m e n t w h e n th e y d e te c t o n l y fiv e
H. armigera caterpillars in a sample
of 25 cotton plants. Even a few of the-se larvae can cauthe-se crop damage — it can be difficult for them to compre-hend the tre a tm e n t c o s t/c ro p loss
Pesticide storage in a cotton extension company warehouse.
Photo CIR AD-U R tA
Treatment decisions
D ecision s to begin treatm ents are made when the threshold has been surpassed. When the scout is using a pegboard, the threshold is surpassed when the peg (inserted successively in the pegboard holes after sampling each u n it) is lo c a te d in the "r e d z o ne ", w h ic h also shows the most s u it a b le f o r m u l a t i o n to be used against each specific pest.
Theoretically, when data sheets are used, trea tm en t thresholds can be adjusted on the basis of the n o m o -graph results, but this w ould require fu rthe r tra in in g on such advanced technical skills.
Formulation and treatment
choices
Staggered-targeted control training in Benin. Photo CIRAD-UREA
trade-off. In this situation, theories c o n c e r n in g b e n e fic ia l fa un a and environment-friendliness should not yet be put forward. This initial step should be handled by local training agencies.
Assessment of infestation levels
The representative sample of 25 cot-ton plants should be chosen random-ly w ith in the field to be protected. The main problem is in predetermi-ning u n ifo rm c ro p fie ld units to be sam pled. This c h o ic e is g en erally made by the training agent through a rough assessment of each field plot (2 ha maximum). Large cropfields are divided up (e.g. in Cameroon, one sample for fo u r 0.25 ha plots in a block design).
The scout then ca re fu lly examines the c o lle c te d sample and assesses infestation levels of the main pests: mites, s u c k in g insects, lea fe atin g c a te r p illa r s and b o l l w o r m s . This agent must be able to recognize all of the life -c y c le stages for each crop pest as well as the damage they cau-se. These counts can be hampered by poor climatic conditions or the vege-tative growth of the cotton plants.
C on trol scouts are responsible for choosing the most suitable pesticide f o r m u la tio n and dosage. In m an y countries, the colours of labels on the pesticide products facilitate choice- m a k in g as th ey co rresp on d to the pegboard or to the pest identification guide. The training staff have to be h ig h ly attentive as there are many possible sources of error.
Spraying techniques vary according to the pest to be controlled. Spraying is directed at the most affected area, i.e. the underside of leaves for mites and sucking insects, plant tops for bollworms, and peripheral leaves for leaf-eating caterpillars.
Training methods
Local extension agencies and young literate scouts are still essential com -ponents of the new pest management programmes. Head farmers have to be trained to ensure the successful e x te n s io n o f these p ro g r a m m e s . SC H W A R T Z (1993) estimated that 13-15% of Burkina cotton producers are literate (8% o f other farmers). D J IM R A O U (1 993) re p o rte d th at 25% of Malian cotton producers can read and w rite Bambara or French. Training methods have to be simple and illustrated, and adapted to this slowly-changing situation.
The tr a i n in g fo c u s o f research is mainly on identification of crop pests and th e d a m a g e th e y ca use . In Cam eroon, ASFOM (1994) re c o m -m e n d e d s e ttin g up a n e t w o r k o f untreated plots to help in tra in in g scouts and field technicians and to monitor pest and disease levels. Extension agencies have access to many tra in in g aids: slides, videos, illustrated leaflets, technical posters, etc. Slide shows have been designed to s h o w h o w V L V tr e a tm e n ts (DE G U IN E & AS FO M , 1990), and staggered-targeted control operations (AMIOT eta!., 1992) are organized. These documents complement prac-tical field sessions, inv o lv in g ento-mologists, aimed at presenting diffe-rent strategies to tr a in in g agents. Scientists provide them with tools to h e lp th e m in u n d e r s ta n d in g pest management principles and gaining skills in this field: pest quantification to determ ine treatm ent thresholds, collection periods, active ingredients and doses. Trained staff are then res-p o n s ib le fo r tr a in in g farm ers and scouts on data collection, treatment d ecision-m aking, c o nd uctin g treat-ments, sprayer handling and mainte-nance.
Research organizations are not yet involved in specific training on pesti-cides: storage, handling, hazards, or their effects on target pests and beneficial organisms. Audiovisual d o c u -m ents d e s ig n e d fo r th is p u rp o s e could enhance toxicant monitoring, reduce pollution and increase treat-ment efficacy.
Conclusion
This new pest management concept addresses econom ic and ecological concerns, i.e. im p ro v in g the cost- effectiveness for farmers while being e n v ir o n m e n t- fr ie n d ly . The role of scientists is to develop pest-specific ecoregion techniques that are adap-ted to different pestmanagement eco- regions, w h ile assisting extension agents in their task of informing and training farmers.
Staggered-targeted control has been w idely used, with promising results. In Cameroon, since 1990, thousands of ha were controlled by this strategy, w ith savings of as much as half the costs required for standard control w h ile still achieving the same crop yields. Benin, Burkina and Mali are currently at the pre-extension stage c o n c e r n in g th is n e w a p p r o a c h . H o w e v e r, there is a higher risk of m a k in g the w r o n g d e c is io n w it h these n ew pest m a n a g e m e n t p r o -grammes. T reatm ent decisions are based on scouts' assessment of infes-tation levels, but crops can be quick-ly and seriousquick-ly damaged once the threshold is surpassed. T raining of scouts, field supervisors, and heads o f fa rm e rs ' o r g a n iz a tio n s s h o u ld involve inform in g and giving them responsabilises, w hile carefully pre-senting them essential g uidelines. This c o u ld be e ffic ie n tly don e by setting up annual 1-2 week training sessions in different countries, w ith an introductory session follow ed by yearly brush-up courses. This training could be partially funded by village associations. Extension of these new techniques should be done gradually in order to evaluate their impacts on the e c o l o g i c a l b a la n c e b e tw e e n pests, beneficial arthropods and host plants.
The fi rst e x te n s io n e x p e r im e n ts re vealed th a t the m ain p ro b le m s encountered are socioeconomic, not t e c h n i c a l , th u s h i g h l i g h t i n g the importance of human aspects of agri-cultural development.
