resource poor farmers

8/4/2019 Resource Poor Farmers

1/12

Plant Disease / Vol. 85 No. 7684

Rebecca NelsonInternational Potato Center (CIP), Lima, Peru, and International

Rice Research Institute (IRRI), Manila, Philippines

Ricardo Orrego and Oscar OrtizInternational Potato Center (CIP), Lima, Peru

Jose Tenorio

CARE-Peru

Christopher MundtInternational Rice Research Institute (IRRI), Manila,Philippines, and Oregon State University (OSU), Corvallis

Marjon FredrixFAO Intercountry Programme on Rice IPM, Vietnam office

Ngo Vinh Vien

National Institute for Plant Protection, Hanoi, Vietnam

Working with Resource-Poor Farmersto Manage Plant Diseases

Farmers in developing countries havesubstantial difficulty in managing plantdiseases (4). Poor farmers understandingof disease processes is limited, and theirdisease management is often ineffective(21). This is, in part, because they cannotsee the organisms that cause plant disease.They often lack access to information andtechnology that could help them raisehealthy crops. Here we present and com-pare experiences in working with farmersto manage rice blast and potato late blight.In these cases, farmer groups learned aboutdisease processes and management tech-niques, and tested promising crop varietiesand breeding lines with the support of ex-tension and research organizations.

The work was conducted using thefarmer field schools (FFS) approach.This involves an intensive, hands-on train-ing program following the extension meth-odology pioneered by the FAOs Inter-country Programme on Rice IPM in Southand Southeast Asia (10,23). Since thisprogram began in the early 1980s, millionsof Asian rice farmers have been trained inIntegrated Pest Management (IPM)through FFS. In a typical FFS, 25 farmersmeet for a weekly half-day session with atrained facilitator over the full course of acropping season. The farmers conduct afield experiment comparing IPM to con-ventional practice, and carry out variousactivities to learn about agroecologicalprinciples. Sessions take place in or nearan experimental field. Each session haslearning objectives and includes groupdynamics to break the ice, hands-on ac-tivities such as development of an insectzoo, evaluation of field experiments, dra-

mas, homework, and evaluation. The ses-sions require careful planning, and thefacilitators role is critical to promote anappropriate environment for research andlearning. Substantial training is needed fora person accustomed to conventional ex-tension to be an effective facilitator.

FFS often focus on management of in-sect pests in rice. Disease managementpresents different challenges than insectmanagement, and different pathosystemsmay present unique problems (Table 1). Inthe case studies described here, field ex-periments were mostly aimed at testingindividual disease management compo-nents. Some of these tests were conductedto demonstrate known phenomena, such asthe increase of rice blast disease severitywith increasing nitrogen input. Other ex-periments were considered farmer partici-patory research (FPR), which was orientedto develop or evaluate technological op-

tions such as the testing of elite breedinglines and genotype mixtures. The approachused is therefore designated FPR-FFS todistinguish it from FFS, which are morepurely oriented to farmer training.

Participatory Researchin International Agriculture

The technology transfer model has beenthe dominant approach for agriculturalinnovation. A top-down approach can beeffective in some cases, as exemplified bycereal breeding efforts aimed at relativelyhomogeneous production systems, and bythe rapid uptake of chemically based tech-

nologies such as pesticides and fertilizers.But in many cases, technologies developedwithout farmer input are not suited tofarmers real or perceived needs and arenot adopted (15). For knowledge-intensivetechnologies such as pest and disease man-agement (here subsumed under the um-brella of IPM), farmers may not be able toutilize technologies without substantialaccess to information and training. To im-plement IPM effectively, farmers mustadapt their management strategies andtactics to local pest complexes, croppingconditions, and operational constraints.

There is an increasing appreciation forthe importance of involving farmers inprocesses aimed at improving agriculturalpractice among agricultural research, ex-

tension, and development organizationsoperating in developing countries. In 1982,Rhoades and Booth (16) argued that in-volving farmers in the research processincreases the chance of success in the gen-eration of appropriate agricultural technol-ogy. Since then, numerous publicationshave documented the advantages of farmerinvolvement in research, extension, anddevelopment efforts (15,17). Participatoryapproaches offer researchers a mechanismto ensure that their work is relevant tofarmers needs and conditions.

The collaboration of farmers and exten-sionists in participatory research may per-mit the collection of substantial datasets,enabling researchers to sample a broader

range of environments than is possiblewith conventional on-station or even on-

farm research. For farmers, collaborationwith the formal research sector offers op-portunities for continuing education oncrop management, as well as early accessto new technologies such as improvedvarieties. For extension organizations,involvement in participatory research of-fers access to innovative problem-solvingapproaches and other types of intellectualcapital. Ortiz (13) indicates that extensionworkers can enrich their cognitive capitaland enhance their decision making processby being involved in participatory research

and training.There are a number of well-developed

models for farmer participatory research.One successful approach is the CIAL (forits Spanish acronym for local agriculturalresearch committee), which was devel-oped by researchers at the InternationalCenter for Tropical Agriculture in Cali,Colombia (3). Over 250 CIALs were ac-tive in the year 2000, mostly in Latin

America and Africa (6). Through CIALsand other approaches, participatory plantbreeding has been shown to be an effectiveway to select locally adapted genotypes

Corresponding author: Rebecca NelsonE-mail: r.nelson@cgiar

Present address of Jose Tenorio: c/o FAO-Lima,Peru.

Present address of Marjon Fredrix: Remalunet29c, 6221 JE Maastricht, Netherlands.

Publication no. D-2001-0423-04F

2001 The American Phytopathological Society


2/12

Plant Disease / July 2001 685

and to improve farmers access to usefulcrop genetic diversity (9,18,19,24).

Improving farmers ability to managecrop diseases requires both knowledge andcapacity for innovation. Without under-standing basic issues such as the pathogenas causal agent, sources of inoculum, andthe concept of latent period, farmers areunable to grasp the basis for disease man-agement strategies. To provide both infor-mation and technological options, we have

embedded a substantial element of partici-patory research into a farmer training pro-gram. This is consistent with the analysisof Braun et al. (6), who suggest that thetraining elements of the FFS approach arecomplementary to the participatory re-search methods of the CIAL, and that itwould be useful to combine elements ofthe two approaches to confront many prob-lems in agricultural production.

Managing Rice Blastin Central Vietnam

We describe here activities that wereconducted in central Vietnam from 1994 to

1997 as a small component of Vietnamsnational IPM program. The work involvedcollaboration among a group of nationaland international organizations, which arelisted in Table 2. The collaboration wasundertaken to help Vietnamese rice farmersimprove their management of rice blast bylinking the strong rice-breeding program atthe International Rice Research Institutewith Vietnams strong rice extension pro-gram, which was supported by the FAOsIntercountry Program on Rice IPM (FAOIPM program).

The Vietnamese IPM program. TheVietnam IPM program was initiated in

August 1992 through a collaboration be-tween national and local agencies, and theFAO IPM program. As of October 1995,the IPM program had trained a total of1,229 Vietnamese extension workersthrough a season-long Training of Trainerscourse. These trainers had conducted 5,000FFS in the country, covering 3,095 of thecountrys 9,300 communes and providingdirect training to 132,125 farmers.

One of the goals of the national IPMprogram was to strengthen the farmergroups that had undergone FFS trainingand to support them in follow-up activities

that could further develop farmers skills inpest and crop management and allow themto tackle additional challenges facing theircommunities. Groups of farmers that hadparticipated in an FFS often declared

themselves to be Farmer Clubs and werekeen to undertake additional training andresearch efforts to pursue possible oppor-tunities or to solve local problems.

Rice blast, caused by Pyricularia grisea(teleomorph: Magnaporthe grisea), wasone of the constraints facing Vietnameserice farmers. The disease was particularlydamaging in central Vietnam, whereweather conditions favor disease develop-ment because the rice crop is grown on arelatively thin strip of irrigated land be-tween the sea and the hills. In spite of theavailability of powerful components thatcould contribute to integrated management

of the disease, farmers were often unableto manage the disease effectively.

Through discussions among staff of theFAO IPM program and the InternationalRice Research Institute (IRRI), it was pro-posed that farmer groups might undertakeFFS activities in central Vietnam. Twocommunities were selected for initial pilotFPR-FFS for the 1994-95 cropping season.Members of these communities, Ha Lamand Duy Xuyen, had previously partici-pated in FFS. They expressed interest inlearning more about rice blast, about dis-ease management, and about testing new

rice varieties and breeding lines. Throughdiscussions with members of the nationaland local extension services, it was agreedthat the FPR-FFS would be facilitated bymembers of the local extension service of

Danang Province, who would providetraining support for weekly training andresearch activities.

FPR-FFS curriculum on rice disease

management. A season-long training pro-gram on management of rice blast diseasewas developed, based on the format andmethodology developed by the FAO IPMprogram. The curriculum was described ina field guide, which was written to serve asa guide for FFS facilitators. The first draftwas based on the ideas of rice pathologistsand extension specialists; this was thentranslated into Vietnamese and adapted andmodified by the FFS facilitators and the

participating farmers. The season-long FFScurriculum involved a set of field experi-ments supplemented by a series of learningactivities. The field experiments includedtesting of promising varieties and breedinglines, testing of four different varietal mix-tures, testing of a range of seed densities,and testing of different nitrogen treatments.The learning activities included discus-sions, observations, manual simulationexercises, and games.

Farmers learned about the nature of hostresistance in several types of activities.Direct field observations were very con-

Table 1. Features of pest management cases, contrasting insects on rice, rice blast in southeast Asia, and potato late blight in the highland tropics

ProblemRole of pesticides in

managementFRole of varietal

resistanceAvailability and efficacy of

other control measuresLevel of farmer

knowledge

Scientists need forfarmer participatoryresearch (subjective)

Insects on rice(SE Asia)

Used but not needed(counterproductive)

In use High (natural enemies) Intermediate Low?

Rice blast(SE Asia)

Used if pressure is highand resistance is low

In use; often notdurable

Medium (agronomic prac-tices)

Low High (deployment ofresistance)

Potato late blight(highland tropics)

Essential in many produc-tion systems, even when

using resistance

In use; often notdurable

Low (agronomic practices;but sanitation ineffective)

Low High (epidemiology anddeployment of resis-

tance)

Table 2. Organizations involved in farmer field schools focusing on participatory researchand farmer training on plant disease management (FPR-FFS) in Vietnam

Organization Role

Farmers groups Conduct field experiments; collect and analyze data; formu-late and design new experiments for local conditions (somecontinued testing varieties for several seasons); share in-formation with other farmers

The National Plant ProtectionDepartment (PPD)

Select farmers groups; assist in developing and implement-ing new exercises for FFS and Training of Trainers (ToT);coordination

FAOs Intercountry Pro-gramme on Rice IPM Assist national program; funds to support farmer groups andtrainers; coordination and documentation

PPSD - IPM trainers Assist in selecting farmers groups and arrangements forfield site; facilitate weekly training and research sessions

The National Institute forPlant Protection

Provide seed and technical support to facilitators

The International Rice Re-search Institute

Draft Field Guide (training curriculum); provide seed andsupport to national researchers


3/12

686 Plant Disease / Vol. 85 No. 7

vincing; resistance was striking, as manyof the entries tested showed no disease.Manual simulation modeling exerciseswere also conducted to show how diseasespreads and to illustrate the effects ofquantitative resistance and of mixture ef-fects (Fig. 1A). For the simulation exer-cises, the farmers drew a grid on a piece ofposter paper. Each square of the grid wastaken to represent a single plant. One ormore beans were placed in random squares

to represent the initial inoculum. For eachinfection cycle, additional beans wereplaced in the eight squares surroundingeach infected plant. Each infected plantwas then marked (for instance, by placinga paper clip in each box with one or morebeans). With the next infection cycle,another eight beans were placed in thesurrounding boxes for each infected plant,and the marker was removed. (Without themarking and unmarking ofplants beforeand after disease spread, the processbecomes hopelessly confusing.)

Variants on this basic idea were used toillustrate various points about disease de-

velopment. The potentially impressivecumulative effects of differing levels ofquantitative resistance were shown bycomparing the results of spreading 2, 4, 6,and 8 beans for each infected plant (Fig.1A compares the results of spreading 4versus 8 beans per plant per cycle). Toshow the potential benefits of using hostgenotype mixtures, different boxes werecolored to represent different types ofqualitative resistance. Different types ofbeans were used to represent differentpathogen races, able to differentially attackdifferent-colored host plants. The results ofthese simulations were impressive to re-

searchers, extensionists, and farmers alike.Farmers also learned about resistance

and virulence through a card game (Fig.1B). For each pair of players, one made

and played a hand of host genotypes, whilehis or her opponent made and played ahand of pathogen genotypes. The formerconsisted of every combination possible ofthree resistance genes, illustrated as pad-locks (red; blue; green; red and blue; redand green; blue and green; red, blue, andgreen), while the latter consisted of everypossible combination of correspondingkeys. As the players pitted a randomlydrawn pathogen card against a randomly

drawn plant card, they grappled with thegene-for-gene concept. By the end of thegame, they realized that a system of threeresistance genes could generate manyresistant genotypes, but that the pathogencould match and overcome these. Thefarmers went beyond the intended scope ofthe exercise by pointing out that theywould prefer to lock up the rooms of ahouse with different locks, so that a ma-rauder with one key could not wipe themout. They made the explicit link to thedesirability of using diverse rice varieties,to avoid the boom-and-bust cycles thatthey had already experienced with blast

resistance.Farmers learned about the environ-

mental conditions favoring blast so theycould evaluate the risk of a major epi-demic. As farmers were already aware ofenvironmental influences on disease devel-opment, this was mostly a case of poolingand reinforcing knowledge. A card gamewas used in which farmers drew a handof conditions and then discussed the ap-propriate management decision to be taken.

FPR-FFS experiments for rice dis-

eases. In its initial pilot version, the fol-lowing treatments were tested in the FPR-FFS: approximately 50 rice genotypes (10

m2 per entry), four genotype mixtures, andsix nitrogen treatments. The designs andtreatments varied somewhat over time andamong the different communities. Each

treatment was unreplicated, as replicationwas sought over communities (5). In retro-spect, we think it would be worthwhile todecrease the number of treatments andinclude at least two replications per treat-ment, to help farmers in making sound

judgments (this was done in the late blightcase study presented below).

Participatory varietal selection. Duringdiscussions held with farmers prior to initi-ating the FFS focusing on rice disease,

farmers indicated that they had only a sin-gle rice variety available to them for theblast-prone winterspring season. Thisvariety, IR17494, was an IRRI breedingline introduced as part of a brown-planthopper testing nursery in 1983. It wasinitially resistant to blast, but this resis-tance had eroded and had been ineffectivesince 1991. The province of Quang NamDanang was not served by any ricebreeder, and national breeding efforts fo-cused on the larger Red River Delta to thenorth and the Mekong Delta to the south.

Farmers were enthusiastic about testingnew varieties and breeding lines. They

were also sensibly conservative aboutadopting them. Farmers wanted data frommultiple sites and years before adopting anew variety, for the same reasons that re-searchers require the same. Although sev-eral of the entries showed promising resis-tance to blast and potentially higher yieldsand better quality than the farmers stan-dard variety, they were not hasty in replac-ing it.

The rice genotypes tested were chosenby the National Institute for Plant Protec-tion and the International Rice ResearchInstitute (IRRI). Vietnamese entries in-cluded genotypes from various national

institutions. In the first season, 42 geno-types were tested in Ha Lam and 46 weretested in Duy Xuyen. The best-performinglines were tested in subsequent seasons,matching crop duration to the needs of theseason. For the 1995 summer season, forwhich short-duration genotypes are de-sired, the farmers in Ha Lam retested the10 best short-duration entries, while thosein Duy Xuyen elected to plant 23 short-duration genotypes and 15 long-durationgenotypes. In later seasons, the number ofvarieties tested in first-season FPR-FFSwas reduced to 12, and a larger plot size(100 m2) was used for each.

Farmers groups and the National Insti-tute for Plant Protection conducted parallelmultiyear tests of the varieties introducedthrough the FFS, and two blast-resistantvarieties were released. The variety MT6,developed by the Food Crops ResearchInstitute, was selected by farmers in HaLam and was subsequently released. MT6is now planted on 10,000 ha in Quang NamProvince, having replaced the susceptibleIR17494 in the most blast-prone of the41,000 ha of rice in that province. Thegenotype 88-65, developed by the VietnamAgricultural Science Institute, was selected

Fig. 1. Tools for teaching concepts of plant disease. A, Results of manual simulationexercise (bean modeling) used to illustrate disease progress for rice blast, comparinga susceptible host (left) with one carrying quantitative resistance (right). At left, eightbeans were placed in boxes surrounding each infected plant (box with a bean) foreach of three infection cycles. At right, four beans were placed in the boxes sur-rounding each infected plant for each of three infection cycles. B, Card game used toillustrate the interaction of plant resistance and pathogen virulence. One player, rep-resenting the host plant, has a set of cards with all possible combinations of locks inthree colors. His or her opponent has a set of cards with all possible combinations ofkeys in three colors. For each turn, each player plays a card at random. If the keysmatch the locks (if the pathogen could attack the host), the pathogen player takes thehand. If any lock is unmatched, the host player takes the hand. Left, card combina-tions in which the pathogen player wins; right, combinations in which the host playerwins.


4/12


by the FFS participants and then releasedunder the name Xi 21 in July of 1996. Xi21 is grown on approximately 50,000 ha innorth and central Vietnam for both winterspring and summer crops.

Rice genotype mixtures. Host genetic di-versity can be useful for suppressing plantdisease epidemics (11,26). With the aim ofidentifying specific cultivar mixtures thatwould be effective and acceptable in cen-tral Vietnam, four genotype mixtures were

tested in the FPR-FFS, each with three tofour component genotypes. Two were con-structed using IRRI breeding lines. A set of10 elite IRRI rice lines was inoculated witha range of Philippine isolates of the riceblast pathogen, and lines showing differen-tial interaction patterns (indicative of theirpossession of distinct resistance genes)were combined. Similarity of grain sizewas also taken into account, to ensure thatthe harvest could be milled together. Theother two mixtures were designed fromrice varieties available in Vietnam, basedon their growth durations to ensure that thecrop could be harvested together.

The results of the farmers experimentswere variable. For instance, in the firstseason, the mixtures performed substan-tially better than expected based on theircomponent genotypes at Duy Xuyen, butthe same mixtures performed substantiallyworse than expected at Ha Lam. In someexperiments, the mixtures were not consid-ered to be sufficiently uniform. Overall,farmers interest in cultivar mixtures wasnot strong.

Nitrogen. Nitrogen is well known tohave strong effects both on rice blast andon rice yield (2). The intensification of riceproduction in northern Vietnam has haddramatic impacts on increasing pests anddiseases, in particular rice blast and sheathblight (caused byRhizoctonia solani) (20).While leaf nitrogen is a strong predictor ofrice yield, levels of applied nitrogen arenot well correlated with yield in surveys offarmers fields because of the wide range

of available nitrogen in rice soils (7). Thus,it would be valuable for farmers to be ableto optimize nitrogen inputs on a very locallevel, taking into account effects on bothyield and blast. Therefore, experimentswere conducted using various levels ofnitrogen, with and without fungicide treat-ment.

As expected, a clear relationship wasobserved between nitrogen input and blastincidence (Fig. 2A). Fungicide treatmentwas apparently not effective. Yield in-creased with nitrogen treatment up to acertain point, and then decreased. To en-sure that farmers understood that different

fields differ in their basal nitrogen levels,farmers experimented with a chlorophyllmeter (SPAD meter; 22) and/or color tabs(25). The color of the leaves indicated thelevel of nitrogen taken up by the plants.With experience, farmers could learn tooptimize their nitrogen level in such a wayas to minimize disease and maximize yield.

Plant density. High plant densities canbe associated with increased fungal disease(1). Farmers had relatively recentlychanged from transplanted rice to directbroadcast seeding, and seed rates in usewere very high. For the first season of

FPR-FFS, small areas with reduced plantdensity were included within the non-sprayed, standard nitrogen plots. The re-sults of these initial experiments indicatedthat lower plant density resulted in ahealthier crop, so the farmers decided toinstall a separate experiment on seed den-sity.

To assess the effect of plant density ondisease and yield, an experiment was con-ducted testing six planting densities. Dis-

ease was clearly affected by plant densityas attained through varying seed rate (Fig.2B). Seed rate also influenced brown spot,caused by Bipolaris oryzae (syn. Helmin-thosporium leaf spot). Again, fungicidewas observed to be ineffective, presumablydue to relatively low disease pressure.Yield was adversely affected by high seedrates, including the rates used by the par-ticipating farmers. Subsequent to theirparticipation in the FPR-FFS, farmersindicated that they reduced their seed ratesubstantially and were saving on seed andfungicide costs as a result, while observinga generally healthier crop.

Fungicide. For each of the nitrogentreatments, half of the plot was sprayedand half was not sprayed with fungicide.At lower levels of disease pressure, fungi-cide effects were not impressive. At ex-tremely high disease pressures, however,fungicide would be important to farmersgrowing susceptible varieties.

Development and expansion of the

FPR-FFS for rice diseases. We recog-nized two phases in the development of theFPR-FFS. Phase I of the program (1994 to1996) involved development, testing, andimprovement of the curriculum based on

Fig. 2. Effect of nitrogen input, A, andseed rate, B, on rice blast severity. A,Data represent average values of unrep-licated plots managed by farmersgroups of Ha Lam and Duy Xuyen vil-lages in the 1994-95 season. B, Datawere obtained from unreplicated plotsby farmer field school group of Binh AnVillage, 1995. F = fungicide, Bl = leafblast, NB = neck blast, F = fungicidetreatment.

Fig. 3. Vietnamese farmers, A, conducting agroecosystem analysis, and B, preparingdata posters after making field observations. Photos: R. Nelson, International RiceResearch Institute (IRRI).


5/12


the experience of the farmers and trainersin central Vietnam. Phase II involved ex-pansion of participation, further refinementof the curriculum, and selection of newresistant varieties. Workshops bringingtogether members of diverse organizations,including experienced facilitators, newtrainers, and researchers, were very impor-tant for development and improvement ofthe curriculum, as well as for training offacilitators (12).

As noted above, two communities incentral Vietnam participated in the firstseason (winterspring crop of 1994-95),and four communities participated in thesecond (summer season of 1995). Photo-graphs in Figures 3 and 4 show aspects ofthe FPR-FFS program. During the firstseason, a severe blast epidemic occurred innorthern Vietnam, which created demandfor participation in other parts of the coun-try. Eighteen communities participated inFPR-FFS for rice diseases during the 1995-96 season, including two communities eachfrom the provinces of TT Hue, Phu Yen,Ha Bac, Nam Ha, Thai Binh, Nghe An, and

Bac Thai, as well as communities fromQuang Nam Danang. In the summer of1996, FPR-FFS focusing on rice diseaseswere initiated in southern Vietnam, wheresheath blight is more of a problem thanrice blast. By the year 2000, 87 FPR-FFSfocusing on disease management had beenconducted with the support of the FAO andthe national IPM program, and another 97had been conducted with the support ofprovincial governments. Thus, a total of atleast 4,500 farmers have participated in thetraining program.

Conclusions: Rice blast management.

Overall, farmers indicated that the FPR-

FFS on disease management were mostworthwhile. The rapid increase in thenumber of farmer groups participating inthe activity and the increasing geographicalrange suggest that this was more than cour-tesy. First and foremost, the farmers wereenthusiastic about having a greater under-standing of plant disease, which had hith-

erto been dangerously mysterious.The farmers showed themselves to be

excellent researchers and decision-makers.It became clear that rice blast could bemanaged, at least for the most part, throughthe use of resistance coupled with opti-mized nitrogen management and optimizedseeding rate. Farmers interviewed indi-cated that, based on their observationsthrough the FFS, they had reduced bothseed and nitrogen inputs, and observed that

their crops were healthier. The farmersreadily adopted experimental methods andused them to improve their disease man-agement strategies.

Blast pressure varied among sites andyears. The performance of any one compo-nent of disease management depended onthe environmental and plant conditions.Because it is dangerous to draw conclu-sions from any one experiment, farmersfound it very useful to conduct exchangevisits to each others fields. They spenttremendous energy on their experiments,which were managed superbly, and tookconsiderable pride in showing their fields

to neighbors and to members of otherfarmer groups.

Deployment of rice varieties turned outto be more challenging than initially imag-ined. Although the local variety IR17494had become very susceptible to blast, itwas still so well adapted to the local envi-ronment that it was often difficult to out-yield. Due to changes in staff at IRRI, theongoing input of new and/or promising ricegenotypes into the farmers experimentswas not aggressively pursued. Thus, al-though FPR-FFS for rice diseases have thepotential to contribute to rice deploymentand diversification, this promise has not

yet been thoroughly fulfilled.

Managing Potato Late Blightin Northern Peru

CARE-CIP collaboration on potato

pest and disease management. Peruviangovernment extension systems related toagricultural pest management were weakor nonexistent in the 1990s. Nongovern-mental organizations were the most impor-tant source of agricultural technology andinformation. From 1994 to 1997, the Inter-national Potato Center (CIP) and the non-governmental organization CARE-Perucollaborated in working with Andean

farmers to improve IPM of the potato crop.During that period, IPM work focused onthe most important insect pests of potato,the Andean potato weevil, Premnotrypesspp., and the potato tuber moths, Symmet-rischema tangolias and Phthorimaea oper-culella. Starting in 1997, CARE and CIPextended the collaboration to include amajor focus on potato late blight.

Late blight was considered to be one ofthe most important constraints to potatoproduction in Peru, as well as in otherdeveloping countries. According to expertestimates, approximately 15% of the Peru-

vian potato crop is lost to late blight, al-though farmers spray fungicides for lateblight an average of six times per season(T. Walker, CIP, personal communication).Every 20% increase in disease severity canresult in a reduction of about 1 t/ha ofyield, which is significant for farmers whoharvest between 5 and 8 t/ha on average(14). Complete loss of the crop is not un-common. Recent changes in the pathogenpopulation had rendered the problem in-

creasingly severe, and metalaxyl, a keyfungicide, was no longer effective (W. G.Perez, J. S. Gamboa, Y. V. Falcon, M.Coca, R. M. Raymundo, and R. J. Nelson,unpublished).

To better understand the context of thepilot effort, CIP and CARE conducted abaseline survey in 1997-98 in the regionwhere FPR-FFS were conducted. Ortiz etal. (14) analyzed farmers perceptions andpractices, and field-level damage in threeprovinces in the department of Cajamarcain northern Peru. This was intended toserve as a baseline study for the FPR-FFS,to allow analysis of the impact of the pro-

ject, as well as to provide informationneeded for design of the FPR-FFS. Thestudy revealed that while farmers derivedincome from a range of agricultural andother activities, potato was their most im-portant crop. More than 90% of the 131farmers surveyed considered late blight tobe an important problem. The majority(67%) rated late blight as their most impor-tant problem, and 24% rated it their secondmost important problem. While the major-ity of farmers were well aware of theweather factors that favor disease devel-opment, few (9%) were aware that it iscaused by a pathogen (14). The majority of

the farmers (88%) were not able to distin-guish late blight lesions from other foliarlesions. Although late blight was clearlyimportant, it was not the only key pestcited by farmers; Andean potato weevil,frost, flea beetles (Epitrix), rots, potatotuber moth, and other pests were also con-sidered destructive.

The majority of farmers interviewed(94.5% of 131 subsistence and semi-commercial small-holders) used fungicidesas their principle method of late blightmanagement, with an average of 6.6 spraysper season (14). Among the 887 sprayapplications made for late blight, the vast

majority (>97%) involved dithiocar-bamate-type fungicides (maneb group:maneb, mancozeb, propineb, metiram),which are classified by the U.S. Environ-mental Protection Agency as a probablehuman carcinogen (8). In 51.3% of thesprays, these contact fungicides were usedalone, while in 47.3% of sprays, contactfungicides were used in combination withproducts that have systemic or translaminareffects. In 42% of cases, farmers appliedmancozeb in combination with metalaxyl.Surveys of pathogen populations, as wellas farmers and researchers observations,

Fig. 4. Community members observingfarmers field trial in central Vietnam.The trial involved testing a set of breed-ing lines and varieties for resistance toblast and for general adaptation andacceptability, as well as testing of geno-type mixtures, seeding rates, and nitro-gen rates. Photo: R. Nelson, Interna-tional Rice Research Institute (IRRI).


6/12


indicate that Peruvian populations of Phy-tophthora infestans had become resistant tometalaxyl (W. G. Perez and R. Nelson,unpublished). Farmers apply fungicideswith backpack sprayers, mostly withoutany protection (R. Nelson, unpublishedobservations).

Most farmers were aware of the exis-tence of late blight resistant potato varie-ties, but few had access to varieties withadequate levels of resistance. Farmers

expressed strong interest in testing resistantvarieties. Most of the farmers (85.5%)mentioned late blight resistance as a desir-able characteristic of a potato variety. Forthree of the seven most commonly avail-able varieties, late blight resistance was themost frequently mentioned reason for pre-ferring the variety. Another importantmanagement tactic was to avoid planting inthe rainy season, with the substantial dis-advantage that in other seasons yields areconstrained by the availability of water.Opinions varied regarding the effects ofvarious agronomic practices on the disease.

At the beginning of this effort, CARE

had been operating in the San Miguel areafor several years and had established awide range of activities related to commu-nity development. Through existing pro-

jects, a team of extension agents was pro-viding agricultural extension services tomany communities. CARE and CIP agreedto develop a pilot activity to adapt andevaluate the FFS as a way to give Andeancommunities improved access to knowl-edge and technology, and to encouragethem to develop their own solutions toagricultural problems.

In contrast to the situation in Vietnam,the concept of FFS was new both to the

extension personnel and to the farmers.CARE staff felt that the methodology wasof interest and would perhaps have rele-vance as a general approach to communitydevelopment. The farmers willingness toparticipate was largely based on the factthat potatoes were important for them andthey perceived that CARE could helpthem. The communities conducting FPR-FFS on potato agreed to meet fortnightlythroughout the cropping season to conductlearning activities and field experiments.During the 1997-98 cropping season, fourcommunities in San Miguel participated inthe season-long FPR-FFS. Two facilitators

were hired by CARE to support the FPR-FFS, with ad hoc financial support pro-vided by CIP.

FPR-FFS curriculum for potato. Theprimary initial objectives of the FPR-FFSeffort were to provide hands-on learningopportunities for groups of farmers inter-ested in learning about late blight man-agement; to give the farmers an opportu-nity to test and select resistant varieties;and to identify and integrate additionalmethods for disease, pest, and crop man-agement. A basic field guide for facilitatorswas drafted to support the participatory

training dimension of the FFS, and toguide the field experiments. As with therice blast curriculum, the potato field guidedescribes group dynamics, learning-by-discovery activities, experiments, observa-tions, and other tools to facilitate learning.

As a first step in the process of curricu-lum development, a series of workshopswas held. The first was an Andean regionalworkshop conducted by CIP in Quito, Ec-uador, to develop awareness about the FFS

approach among research and extensionorganizations. National workshops werethen held in Ecuador, Peru, and Bolivia,and FPR-FFS were initiated in all threecountries through the interaction of therespective national programs and CIP staffbased in each country. Curricula were de-veloped independently in the three coun-tries. In 1999, a project was funded by theInternational Fund for Agricultural Devel-opment, which supported the developmentof FPR-FFS for potato in six countries(Bangladesh, Bolivia, China, Ethiopia,Peru, and Uganda).

The focus here is on the work conducted

in Peru. In drafting the first version of thePeruvian Field Guide, it was surprising tofind how few of the activities could beeasily transferred from rice blast to lateblight. While the manual simulation gameswere extensively used for the rice blastwork in Vietnam, these activities werepoorly received in the Andes. One reasonfor this is that rice blast showed clearlyfocal epidemics, so that the simulation ofdisease spread was easily related to theprocess occurring in the field. Althoughpotato late blight epidemics involve initialpatchiness and spread in temperate areas,the disease often appears to begin with a

uniform initial level of disease under An-dean conditions, where year-round potatoproduction and the presence of alternatehosts may lead to the presence of highlevels of initial inoculum.

For the potato field guide, the use ofmini-microscopes and moist chambers forthe culture of lesions were important learn-ing tools (Figs. 5 and 6). Farmers usedsmall field microscopes to closely observedifferent types of lesions, and to see spo-rangia of P. infestans for the first time.Many farmers were excited to see the lit-

tle lemons that cause so much destruction,and to realize that these tiny structuresmove through the air and also through thesoil to attack the tubers. The curriculuminvolves a series of small experiments inwhich a farmer places plant tissue in adisposable plastic box or a plastic bag,together with moist tissue paper and asupport (a few twigs) to keep the leaf,stem, or tuber out of direct contact with thepaper. The moist chamber activity is used

first as an aid to diagnosis. Farmers placedifferent types of lesions in the chamberand watch what happens over several days,to learn the difference in the life cycle ofa late blight lesion compared with a lesionincited by Alternaria or a spot caused byinsect damage. Unlike other types of spots,late blight lesions rapidly overcome a po-tato leaf of a susceptible variety.

Farmers then use the moist chamberformat to confirm that the sporangiawashed from a sporulating late blight le-sion do indeed incite new lesions on unin-fected leaves, and that transmission canalso occur when an infected leaf is incu-

bated in the chamber with a healthy leaf.The transmission studies also show farmersthat late blight lesions on leaves can lead totuber blight. While most farmers partici-pating in the FPR-FFS were familiar withlate blight lesions on tubers, they were notaware that the problem is related to thephenomenon of leaf blight. This informa-tion allowed farmers to understand howhigh hilling can keep the tubers safer fromthe pathogen, and how removal of infectedvines can decrease infection of tubers atharvest. The moist chamber can also beused to demonstrate different levels ofresistance and the efficacy and nature of

different types of fungicides.While the Vietnamese rice farmers had

already participated in FFS prior to under-taking a specialized program on rice dis-ease, the FPR-FFS for Peruvian potatofarmers was a unique opportunity to pro-vide access to general information on cropand pest management. Thus, the curricu-lum was designed to cover important is-sues beyond late blight management, andits coverage has broadened steadily over its4 years of evolution. Seed quality is a keyissue in potato cultivation. In the FPR-

Fig. 5. Farmer drawing sporangia ofPhytophthora infestansobserved with amini-microscope. Photo: R. Orrego,International Potato Center (CIP).

Fig. 6. Farmers setting up an experimenton the etiology of tuber blight, Peru.Photo: J. Llontop, International PotatoCenter (CIP).


7/12


FFS, farmers sort seed-lots into differenttypes and discuss and then observe theeffect of seed quality on the health of thecrop. They learn how to maintain seedquality over seasons. In addition, increas-ing emphasis has been placed on manage-ment of key insect pests.

The field guide was initially designed togive session-by-session guidance for theFPR-FFS facilitators. After 2 years of cur-riculum development and modification

with input from facilitators and farmers,the content became too broad to fit into asingle season, and it became clear that it isimportant to tailor the content of the FPF-FFS to the specific needs of each partici-pating community. Therefore, the curricu-lum was reorganized into sections, cover-ing the different types of resources neededby a facilitator: general information, guid-ance for planning a season and a session,group dynamics, field experiments, learn-ing activities, monitoring and evaluation,and technical information. Notebook andCD versions were developed to make thisinformation more accessible to practicing

or potential FPR-FFS facilitators. Theresponsibility for selecting among theseresources to create a useful learning andresearch program was more explicitlyplaced on the facilitator.

FPR-FFS experiments for potato. Par-ticipatory experiments (on-farm trials con-ducted with active farmer involvement) arean important part of the FPR-FFS. For thefirst 2 years, participating communitiestested a set of available genotypes, eachwith three levels of fungicide treatment.During the third cropping season, each FFSgroup conducted four experiments. Oneexperiment compared three potato varieties

with different levels of resistance, eachmanaged with three different fungicidestrategies. Another experiment involved a

set of new clonal breeding lines, and athird involved entries derived from truepotato seed. The final experiment com-pared IPM with farmers practicetreatments; this is the type of experimenttypically conducted in a farmer fieldschool. After the first 3 years, it was con-cluded that these experiments were overlycomplex, and simpler designs are beingimplemented for the fourth season of FPR-FFS.

In a traditional FFS, farmers conductagroecosystem analysis, quantitativelyevaluating the factors that affect the crop,and capture their observations on a largeposter. They typically draw the plant in themiddle of the poster, illustrating the pestsand noting their numbers per plant on oneside, while noting the natural enemies onthe other. The plant condition, as well asthe weather and soil factors, are duly in-cluded and annotated in the drawing. Thisapproach was adapted to allow the farmersto capture the observations made for eachof their experiments. The process of col-lecting data and making posters allowed

farmers to follow the progress of eachexperiment, and later to share their resultswith other groups at field days, and toparticipate in planning and decision-making processes between seasons.

Participatory varietal evaluation and

participatory breeding. During a nationalworkshop held in 1997, a set of promisingpotato varieties and breeding lines wasselected by representatives of the nationalagricultural research institution, nationaluniversities, CARE, and other key stake-holders. For the first 2 years, one ratherlarge and complex field experiment wasconducted. This involved 12 to 14 varieties

and breeding lines identified by breedersand others at CIP and in the national re-search program (INIA and universities).

Each of the varieties was treated with threedifferent levels of fungicide, and eachtreatment was replicated twice.

Data were obtained for 16 varieties andadvanced lines over the three field seasonsfrom 1997 to 1999, with 20 to 90 evalua-tions of yield and AUDPC per entry over-all. The formal analysis of these data willbe presented elsewhere. The weather wasgenerally conducive to late blight, and thesusceptible genotypes were often killed by

late blight irrespective of fungicide treat-ment. The yields of the moderately resis-tant genotypes varied according to diseaseintensity (which varied with local envi-ronment and fungicide treatment), andyields for most of the moderately resistantto susceptible entries were negatively cor-related with mean disease levels. The resis-tant varieties performed well even at lowfungicide levels. The overall mean yield ofthe most resistant entry was nine timeshigher than the overall mean yield of themost susceptible entry. Farmers who par-ticipated in the first years FFS continuedtesting the best clones in their own fields.

The FPR-FFS have provided an oppor-tunity for farmers to gain access to varie-ties previously little known in their areas,as well as to breeding lines not yet for-mally released. The most promising entryidentified through the first years FFS wasAmarilis, a blight-resistant variety withwhich few local farmers were previouslyfamiliar (Fig. 7). After its success in theFPR-FFS, CARE provided credit to allowlarger-scale production of Amarilis, and ithas rapidly gained acceptance in the area.For example, 35% of families who partici-pated in FFS had commercial plots withAmarilis after two cropping seasons of

participation, compared with 10% of non-participant families. This 10% suggeststhat if an FFS group finds a good variety orclone, the material will be made availableto other members of the community.

Starting in 1997, an agricultural coop-erative and the national agricultural re-search institute each provided two clones.One clone per pair was considered to begood while the other was rejected due tolate blight susceptibility, low yield, andtuber deformation. The two good cloneswere formally released by their respectivenominating institutions, becoming nation-ally recognized and available after many

years on the shelf.Based on the 1999 evaluations of eight

FPR-FFS groups, one clone was consis-tently preferred. This clone, Chata Roja,was selected by the National AgrarianResearch Institute (INIA) from a CIP crossmade in 1984 (Fig. 8). In part based on thepositive evaluation of the FPR-FFS, thisclone was released by the Hermilio Valdi-zan National University (Huanuco, Peru)in 2000. Farmers data from three FPR-FFS are shown in Figure 9.

In 1999-2000, a set of 50 newly identi-fied breeding lines was tested among 19

Fig. 7. Farmers enthusiastically observing the yield of a plant of variety Amarilis in thevillage of Lanchepampa in northern Peru. Photo: R. Orrego, International Potato Cen-ter (CIP).


8/12


FPR-FFS groups. Most breeding lineswere tested in three to four communities,and each community tested approximately10 new clones. At the end of the season,the farmers, facilitators, and researchersconducted a workshop to analyze the re-sults of the testing of the new set of breed-ing lines. A subset of the clones was se-lected for continuing evaluation among thecommunities. Each community also testeda set of populations developed from true

potato seed (TPS). As part of the FPR-FFScurriculum, farmers learned about TPStechnology as an alternative approach toseed potato production.

Genotype by management by environ-

ment interactions. In the first and secondyears of FPR-FFS experiments, all of theentries were treated with three levels offungicide protection. In the third season, asmaller experiment was conducted to dem-onstrate the interaction of three genotypeswith different levels of resistance withthree levels of fungicide treatment. For thefarmers, the purpose of this experimentwas to learn that fungicide spray decisions

should be predicated both on the level ofresistance of the crop and on the environ-mental conditions. Analysis of the 3-yeardataset revealed that the effects of differentresistance levels, fungicide spray intervals,and environments on disease were allhighly significant, as were their two-wayinteractions. The three-way interaction washighly significant for disease severity (P

resource poor farmers

Documents