Strategies for the Future - Princeton

Chapter 3

Strategies for the Future

CONTENTS

Page

WHAT USAID DID WELL ● ****4*****.**.******+**”.***”***********”****** 69

Promoting Internal and External Coordination , . . . . . . . . . . . . . . . . . . . . . . . . @ProvidingTraining . . . . . . . . . . . . . . . . . . . .. . .=......” ,.*”.”””~*ooos TOAdvocatingSoundInsecticide Use. . . . . . . . . . . .. .. . . .....=.=. .==. ””-”” TO

HOW TO DO BETI’ER NEXT TIME ● *e ● ● ● * 9 ● ● ● ● 0 ● ● ● * ● * ● e ● ● ● ● ● ● ● ****** ● ● ● ● O 71

The FeasililityandPrice ofPrevention . . . . . . . . . . . . . . . . . . . . . . .....=..” 72Integrating13mergeriq”COntxOl Programs Into ~ng TerrnDeVelOPrnent . . ’731ndividualorMultipestStrategies? . . . . . . . . . . . . . . . . . .. = ... = .......=” ~~Whenand Where ShouldContrOl Programs Remounted? ..*...* . . . . . . .

WHAT CONTRALTO USE: THE ROLE OF TECHNOLOGY . . . . . . . . . . . . . . . . . 75IntegratedPest Management . . . . . . . . . . . . . . . . . . . . . . . . . .....=.-. ...=”. 76MonitoringInsects, Weather, andVegetation . . . . . . . . . . . . . . . . . . . . . . . . . . 81

Figure

3-1 PrincipalSatelMesUs edinEarly Warning andForecasting . . . . . . . . . . . . . . . . 84

Tables

3-1 OTASurveyRespondents: Percent occurrent and Ideal LoCustEffortSFoeused onCrisis,Reliefi andPrevention. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74

3-2 ControlTactics NowEmployed Against Major Pestsof WheatintheU.S. Great PlainsandSorghum inTexas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77

Chapter 3

Strategies for the Future

OTA’S analysis found that the 1986-89 locustand grasshopper control campaigns in Africa werebased on questionable premises, with partially ef-fective to ineffective im lamentation. Yet, somethings worked well and 8.S. efforts contributed tothese successes.

WHAT USAID DID WELL

Finding: US~D & commendable attempts:1) to coordinate its eflorts with other U.S. agenciks,foreign donom, and A&can o~iak; 2) to providetraining for A~ans and U.S. pemonnel; and 3) tohighlight iksues of sound iksectkidk choke, stkrage,applicatwn, and dikposal. Overall, the internathudcontrol campaign lkked these characteristics, how-ever. USAID didprevail successfidly against the useof dieldrin.

Promoting Internal and ExternalCoordination

The U.S. Agency for International Develo -rment (USAID) coordinated its work successfu Iy

within USAID and with other U.S. Governmentagencies involved in the campaigns despite for-midable institutional constraints. The DesertLocust Task Force, established within USAID’sOffice of Foreign Disaster Assistance (OFDA),was one of the most effective means of coordina-tion within the U.S. Government. From July 1988through June 1989, the Task Force held weeklymeetin s to share information, assign res~on-

F’sibility or implementing activities, and coordinateefforts.

Also, OFDA brou ht together people repre-hsentingavariety of US Departments and other

organizations to review results from the reviousfyear’s efforts, to identi~ lessons Iearne , and to

plan more effective future control. OFDA spon-sored two workshops for Task Force membersfrom Washington, DC, USAID mission staff fromAfrica, and outside experts. First, the U.S. ForestService’s Disaster Assistance Support Programmanaged a 3-day workshop in January 1988 inHarpers Ferry, West Virginia, for 69 officials,mostly from the U.S. Government, to evaluate the1986 and 1987 campaigns and provide direction fora staff guidebook on locust and grasshopper

programs. Then, 32 participants took part in a4-day, February 1989, workshop in Dakar,Senegal; they reviewed each country’s 1988 cam-paign and were introduced to the finalized USAIDguidebook.

This 1989 Locust/Grasshop er Management$Operations Guidebook is wel -prepared and

thorough, for the most part. It promdes a com-prehensive overview of USAID’s policies regard-ing locust and grasshopper control, includes usefulback round information on the insects’ biology

iand ehavior, sets forth the rationale and proce-dures for mounting a control operation, providesdetails on conducting insect surveys and selectingappropriate control techniques, and includes help-fulsupplementary information e.g., pesticide-use

$guidelines, procurement proce ures).

OTA expects that the Guidebook will con-tribute to a more expert, consistent, and coor-dinated U.S. response to rasshopper and locust

fproblems in the future. f used effectively, theGuidebook could achieve its purpose: “... to assistMissions to assess, prepare for, and organizelocust/grasshopper control programs on an emer-gency and non-emergency basis” (118, p. I-2).

The Guidebook is the most up-to-date opera-tional source for selecting insecticides for U.S.-funded work and lists a number of selectionconsiderations. However, the database on insec-ticides constantl changes. For example, the U.N.

dWorld Health rganization’s Hazard Classifica-tion, revised every 2 years, now has differentratings for a proximate one-fourth of the es-

Z i &ticides inclu ed in the 19 9Guidebook. US .preparing Country Supplemental Environment~~Assessments in 1990, with technical assistancefrom the U.S. Environmental Protection Agency(EPA), to apply thecontinent-wide ProgrammaticEnvironmental Assessment to the individualcountries planning to use insecticides againstgrasshoppers and locusts. This process, which aimsto make more site-specific plans, could allow up-dated information on different chemical productsto be incorporated in the supplemental assess-ments simultaneously. However, these sup-plemental assessments also will need to be revisedperiodically to remain current.

69

70 ● A Plague of Locusts

USAID actively promoted coordinationamong other donors and African governments,and agreement exists that coordination and col-laboration among countries increased as therecent campai ns progressed. For example, rep-

!resentatives o USAID or the U.S. De artment ofrAgriculture’s (USDA) Animal and P ant Health

Inspection Service attended erhaps a dozen5meetings sponsored the U.N. oodandAgricul-

8ture Organization ( AO) to share informationand plan future strate . USAID funded FAO’S

YEmer enc Centre or Locust Operations(ECL&), tte worldwide coordination site forlocust and rassho per control operations, and

! “cl’USAID sta f prow ed ECLO with data on insect

Fovulations and U.S. control efforts. The Bureauor Science and Technology participated in the

World Bank’s Special Program for AfricanAgricultural Research on locusts.

USAID required that recipient countries havean operational Country Coordinating Committee,composed of representatives from relevantgovernment and donor or animations, before U.S.

femergen~ funds were re eased. USAID missionstaff participated in these committees and alsomaintained direct contact with the national cropprotection setices and other African agencies in-volved in control.

Providing Training

USAIDprovided trainin foritsownpersonnel#and Afi-ican officials throu workshops and the

provision of technical assistance. Additionally, theUnited States funded trainin programs for

dAfricans, conducted by FA and regionalorganizations. For example, FAO trained Saheliannational crop

1rotectlon personnel in locust

surveillance an another group, Application ofAgrometeorology and Hydrology for the Sahel

[TAGRHYME , conducted an annual short course

or African officials on using “greenness maps.” Thistraining and technical assistance, together with theprovision of equipment and supplies, undoubtablestrengthened the ca acity of national institutions tomount future locusJfasshop~rsumyandcontrolprograms and to eal wlt other agriculturalproblems.

USAID conducted 10 training workshopsfrom 1987 through late 1989 with a total of ap-proximately 150 participants. One early workshop

on how to plan and manage aerials rayin opera-tions was attended by Africans ~om &negal,Gambia, Niger, and Sudan. From April throughJune 1989, three regional workshops were held on:1) aerial and ground ultra-low volume (ULV) ap-plication, 2) training extension workers to use newteaching materials on pesticide use, and 3) humanhealth impacts of pesticide application (121). AFebruary 1990 conference on pesticide disposal,held in Niame , Niger, attracted 58 participants

Kffrom 15 West rican countries and internationalorganizations such as Earthwatch and Green-peace. Action plans were drawn up for eachcount . Otherworkshops planned for 1990 areon

?identi lcation of immature Sahelian grasshoppersand crop loss assessment.

USAID developed some useful materials forits training efforts. For example, the PesticideUsers Guuie, prepared in four languages forAfrican extensmn agents, details how to conductpest sumeys, plan insecticide applications, andapply, transport, store, and dispose of pesticides.In addition, USAID funded publication of a fieldmanual for identifying immature grasshoppers(51).

USAID attem ted to increase its own tech-Enical capacity by orrowing experts from other

U.S. agencies and hiring consultants from univer-sities and rivate firms. An effort was made to pair

fsenior an junior entomologists on technical assis-tance teams to increase the pool of expertise avail-able in the future. USAID encouragedparticipation of African officials on the severaldozen U.S. technical assistance teams sent toAfrica. This practice imparts on-the-job training–for those U.S. scientists unfamiliar with Micanconditions as well as for African experts unfamiliarwith some recent pest management technologies.

Advocating Sound Insecticide Use

USAID advocated safe and sound insecticideuse throughout the 1986-89 campaign and en-forced its relevant environmental policies. Itsgreatest success was persuading other donors andAfrican governments not to use dieldrin, eventhough many African countries had existingdieldrin stocks and FAO and France urged its use.With encouragement from USAID, FAO is takinginventory of existing stocks of dieldrin, beginninga study of potential environmental risks of dieldrin

Ch. 3–Strategiesfor the Futire ● 7 1

use in areas where the Desert Locust is present,and intends to develop a plan for use or destruc-tion of dieldrin based on these findings (104).USAID, too, has compiled some information onstocks of dieldrin (99) and sent EPA repre-sentatives to advise African officials on storageand disposal of surpluses.

USAID’s efforts also increased awareness inAfrica of the potential dangers of the persistentorganochlorines and helped reduce the use of ben-zene hexachloride (BHC) and lindane. USAIDencouraged the use of less toxic chemicals and, toa limited extent, tested new insecticides for locustand grasshopper control under African conditions.

USAID promoted increased efficiency insome spray operations, for example, by pro-positioning insecticides in Africa to reduce high airfreight costs. Bysu porting application of satellite

rremote sensing to ocust surveillance and fundingresearch on alternative control methods, USAIDbegan to lay the groundwork for reduced relianceon spraying as the only available response to locustand grasshopper upsurges.

USAID included safety concerns in its techni-cal assistance and training rograms, e.g., by

i!’providing protective clothing or spray operators.USAID claims it was the first to introducecholinesterase testing into locust control programsin Africa. Moroccan applicators were testedbefore, during, andaftersprayin in 1988and 1989

ito determine if the enzyme c olinesterase hadbeen suppressed by pesticides (51).

Also, USAID exhibited concern about the en-vironmental effects of control programs, in par-ticular by preparin$ environmental assessmentsfor Morocco, Tunisia, and all of Africa and Asiaaffected. Since mid-1989, USAID has been design-ing ways to implement the 38 recommendations ofthe Programmatic Environmental Assessment

E . Technical assistance teams are assistingHp. )rlcan nations on the safe disposal of empty con-

tainers and surplus insecticides now thatwidespread spraying is unnecessary.

USAID is seen as among the strictest donorsregarding safe pesticide disposal and is planning totake stronger measures in the future. Its opera-tional Guidebook contains directions for storing,

packaging, labeling anddisposingof pesticides andemp~ containers. An annex contains a co of

PFAOs 1985 Guidelines for the Disposal of astePestide and Pesticide Containerson the Farm thatdetails physical, chemical, and biological disposalmethods. Some other donors have similar inter-ests and a recent workshop on disposal of obsoletepesticides and empty containers in Niameydemonstrated African concern as well.

In short, USAID succeeded in almosteliminating the use of the most hazardous chemi-cal, dieldrm, and identified some lessons learnedfor improved strate ies and tactics for future

Fprograms. The overa 1 locust campaign, however,demonstrated the need for more coordinated ac-tion, far more training, better understanding oflocust and grasshopper d amics and effects on

Pcrop yields, and improve control methods. Forexample, the new Locust/Grasshopper Manage-ment Operdars Guidebook fails to discuss thedebate over the relative roles of control in insectdeclines; USAID’s 1988 training sessions weresidelined when its resources were redirected tospraying activities; USAID’S training and technicalassistance reached only a few Africans; and, insome cases, USAID did not convince Africans ofless toxic chemicals’ effectiveness.

Admittedly, USAID is only one im rtantractor, having provided about one-fifth o donor

funding for recent control campaigns. Thus,USAID has limited responsibility for the failuresof recent campaigns, as well as their successes.

HOW TO DO BE’ITER NEXT TIME

Finding: tinom and~tin governme~ can-not aflord to fund expensive control campaignswithout &mssing@ndame& questions regardinggoals and implementation. Now ii the time tajhdmethodk that wntribute to kmg-tlnn development,redouble pmve~”ve eforts, and decidk WW actiimswill be most efiective during the ti upsurge.

Doing better in the future, during recessionsand upsurges of these insects, revolves areexamination of fundamental questions regard-ing who should do what, and when, where, how,and why it should be done. These are broad policyquestions encompassing all as~cts of control~rograms. For example, which resects should beincluded in programs (individual pests or groups


of similar pests), where control should be mounted(“strategic” areas, breeding sites, or anywhere),when control should be undertaken (when aplague threatens, when swarms threaten crops, orwhenever insects become gregarious), why controlis needed (e.g., to stop plagues, save crops, orprevent famine) and how control is best done (e.g.,aerial or round spraying, four- or single-engine

iplanes or elicopters).

Control or animations, host governments, and

‘onop ‘a v e ‘ie ‘eTnsibili? ‘0 a n s w e r ‘hesequestions. Here, O A ldentl ES some elementsof the discussion and notes that resolution of theseissues shouldbe attem tednowthat upsurges havesubsided for a time. &e roles of various groups–who should do what–also need to be clarified. Thisquestion is addressed in chapter 4.

Further discussion and clarification are espe-cially needed regarding the goals of the controlprograms and indicators to measure their resultswithin specified times. Do the programs aim toprevent plagues, stop plagues, protect cro s, or

?end famine? Different oals imply dif erenttstrategies, action plans, an evaluation criteria.

The Feasibility and Price of Prevention

The FAO and USAID officials responsible forgrasshopper and locust control pro~rams maintainthat knowledge is available that, lf properly ap-plied, could prevent future lagues of locusts andgrasshoppers (12, 95, 121). ~lagueprevention hasconsisted, since the 1960s, of making surveys inseasonal breeding areas and controlling any al-ready-gregarious resects or populations becominggregarious (70). Certainly, the feasibility ofprevention steadily increases as additionalcountries agree to participate in such an approachdurin recessions; as breeding areas are more

fclear y identified; as improved methods aredeveloped for forecasting the rise and movementof insect populations, weather systems, and plantcover; and as more effective, carefully aimed con-trol operations are mounted. However, some fac-tors that contribute to plagues are unresolvable byexisting technologies or largely beyond the controlof donors. These constraints include the un re-

Xdictability of weather and disputes within an be-tween countries. Also, wide-scale implementationof what is known, e.g., about effective spraying, isoften exceedingly difficult under actual condi-

tions. Thus, OTA questions whether donors andaffected countries can prevent upsur es and

~plagues, although that goal is lauda le anddeserves to be foremost.

FAO finds that:

although there is a rational strategy for thep~evention of desert locust pla ues, and tactics and

Jtechniques have been evolv to implement thatstrategy, circumstances can still combine to lead tothe threat of the development of a new major plague.Furthermore such combinations of circumstances,and in particular sequences of widespread heavyrain, cannot yet be forecast

and concluded that:

local outbreaks capable of !eading to major~~urges are likely to be a recurrent but intermittentfeature of Desert must population dynamics. . .(81, cited in 13).

The preventive strategy FAO and USAID ad-vocate thus requires a certain amount of continu-ing monitoring and control. Usuall , that has not

6been done between upsurges. FA and USAIDofficials are requesting funds for a plying this

~strategy now with the ex licit o jective of[preventing future outbreaks romdevelopinginto

plagues.

They, like others, assume that plague preven-tion costs less than plague control. This seemscorrect intuitively but it has yet to be proven.Donor costs of the 1986-89 control campaign,

8rincipally against the Desert Locust andenegalese Grasshopper, were $275 million. In

1988, representatives from several governmentsmet in Fez, Morocco and approved plans for amultinational ongoing survey and control opera-tion to monitor the Desert Locust in its remoteSahelian breedin areas. This International

fDesert Locust Tas Force, with 5 main units and13 sub-units in strategic areas, carried a $77.4 mil-lion price tag. As the plague subsided, the estimatefor Phase I in 1989 was revised down to $3.5 million(106). Thus, the cost of maintaining these mobileunits is far less than the cost of the recent controlcampai n in an equivalent period. However, the

Fcosts o plague prevention v. control should becalculated over a longer time period from abroader base, e.g., perhaps including costs formonitoring and controlling other grasshoppersand locusts and the related expenses of the nation-al crop protection services.

Ch. 3-Strategies forthe~utire ● 73

FAO proposed recently a 5-year re ional!preventive Desert bcust control rogram or the

J8 countries of Maghreb and the ahel. FAO as-serts that control measures in a eneralized in-

Evasion would cost, in 1 year, w at preventivecontrol activities would cost in 15 to 20 years.FAO anticipates that this reventive program

1would cost $6 million to 8 million per year(108,109) and result in less insecticide use over asmaller area, e.g., 50,000 to 100,000 ha per yearsprayed compared to the 15 million ha treated in1987/88 (108). The availability of funding for sucha broad international program has not yet beendetermined. Even if the preventive approachesadvocated by FAO, USAID, and other officialswere fully funded, it seems likely that emergencyefforts would still be needed when the insectsescape strategic control efforts.

Shifting to a preventive approach first requiresa reorientation of thinking by African and donorpolicymakers, followed by corresponding changesin programs and financing. Crises mobihze atten-tion and resources: emergency locust andgrasshopper programs garner far more policy in-terest than long-term efforts, such as inte rated

Fpest management (1PM). Africans favored aster-act-ing insecticides. Emergency spraying opera-tions fit within what some find is a “cowboy”mentality among U.S. officials: a tendency topromote large interventions and uick solutions.

1For example, U.S. officials emp asized use offour-engine planes while FAO and other donorspreferred smaller planes. Thus, preventive ap-proachespresent psycholo icalaswell as technical

fchallenges and their imp ementation would re-quire attitudinal shifts and technical trainingwithin USAID, among other donors, withinAfrican countries, and in Congress.

Integrating Emer ency Control Programs+Into Long erm Development

Donor groups often classify their activities asrelief or development focussed. Generally, reliefactivities are short-term and address symptoms orconsequences of deeply rooted problems. Theycan include actual control efforts and other ac-tivities to help eople recover from losses, e.g.,providing foo c1’ to areas where locusts havedestroyed crops, or providing seeds for replanting.Some also describe activities that help recipientsrecover from control programs (e.g., destruction

of pesticide containers, disposal of surplus stocks,testin operators for over-exposure to insec-

5ticides as “relief and rehabilitation.” Develop-ment activities, in contrast, tend to deal with theunderlying causes of problems and are necessarilylonger term. For example, entomological researc 1to develop safer or more effective control methodsand efforts to prevent locust or grasshopper up-surges would be development actwities.

Individuals and organizations generally con-centrate their efforts on one approach or the otherbecause of the difficulties of combining the two.Some relief efforts incorporate development ob-jectives better than others: e.g., roviding seeds

Frather than food aid, and training armer brigadesto conduct local survey and control programsrather than replacing local efforts with expatriate-run operations. Some relief programs can hamperdevelopment efforts. For example, food aid haslong been criticized as lessening incentives forsmall farmer production although this is not alwaysthe case.

The U.S. foreign assistance mandate encom-

basses both relief and development programs.owever, the recent grasshop er and locust con-

[trol programs seem overwei~ ted by short-termemergency responses despite the well-knownweaknesses of crisis management. Nearly all U.S.funds for locust and grasshopper programs in fiscalyears 1986 and 1987 were OFDAfunds (table 1-3)and 58 ercent of the Africa Emergency

8Locust/ rasshopper Assistance (AELGA)project’s budget for fiscal years 1988 through 1990was allocated to emergency assistance (chemicals,equipment, and short-term technical assistance) v.42 percent for development assistance (research,training, and institutional support) (99). Respon-dents to OTA’S survey agreed that crisis manage-ment (e.g., sprayin programs) was the major type

Eof activity underta en in recent campaigns (table3-l). Most noted the need for a decrease in crisismanagement per se and an increase in bothpreventive measures and specific types of relief,althou h they did not advocate decreasing the

foveral total amount of resources (10). Theiranalysis agrees with that of others (e.g., 95).

The farmers and herders who are the intendedbeneficiaries of donors’ programs do not distin-guish between crisis management, subsequentrelief activities, and long-term development assis-


Table 3-1-OTA Survey Respondents: Percent of Current and Ideal Locust Efforts Focusedon Crisis, Relief, and Prevention

(N = 25)

Current effort Ideal effortMedian (Range) Median (Range)

Crisis 9070 (25 - 100%) 50% (O - 80%)

Relief 5% (o - 30%) 10% (o - 50%)

Prevention 170 (O - 32%) 30% (5 - 100%)

SOURCE: Dale G. Bottrell. ‘locusts in Africa and the Middle East: Summary of Response% to OTA Questionnaire,” contractor reportprepared for the Office of Technology Assessment, May 1989.

tance. For them, locusts and grassho pers repre-l’sent one more crisis in lives that are uI1 of crises,

each further narrowing their options and con-tributing to the downward spiral of poverty (20).Likewise, locusts and grasshoppers are only two ofmany types of pests that threaten their crops. Forlong-term development to succeed, it seems thatfar more attention must be aid to how pest

1’problems interact with otherdi ficulties and to thedevelopment implications of grasshopper andlocust control.

In this context, plant protection needs to beviewed as a process that integrates local, national,regional, and international components. Manyfarmers and herders have few options for control-ling large upsurges of locusts and grasshopperswhen prevention fails. They may need assistanceduring that difficult, but brief, period in which theirlosses can be severe. Thus, short-term relief maybe needed locally, either to prevent crop damageor to enable farmers to recover from that damage,preferably in forms that contribute to long-termdevelopment.

Individual or Multipest Strategies?

General agreement exists that sustainableprotection of crops and livestock requires com-prehensive, multipest management solutions. But,some do not agree that management strategies forlocusts and grasshoppers should be integrated intomultipest management schemes of single organiza-tions, such as the national crop protection services.They note that certain insects require distinctly dif-

ferent control efforts by actors at different levels.Some species, e.g., the Senegalese grasshopperand African Migratory Locust, breed in areaswhere dryland farming predominates and can bemonitored by farmer committees and integratedinto multipest management by the national cropprotection services and farmers. Generally thisapproach could apply to most grasshop rs. On

rthe other hand, species such as the Re Locust,Brown Locust, and especially the Desert Locust,breed in remote areas and migrate across boun-daries. They may be more effectively dealt with asindividual species based on interstate or regionalcooperation. Pro osals are now being consideredfor a regional aJ hoc task force to control theDesert Locust in “strategic” areas outside of WestAfrica’s croplands. The same role was proposedfor the regional organization DLCO-EA in East-ern Africa.

However, addressing locust and grasshopperproblems within thecontextof broader pest problemswould have several advantages: costs would droprelative to benefits because benefits would accrueeach year rather than sporadically; institutional con-tinuity and expertise would be built; already-exist-ing organizations could respond more qulckl to

i?outbreaks and they could accommodate shi tingpest problems methodically; pesticides could beturned over and replenished more rapidly so lesswaste would occur (95). The constraints to adopt-ing a multi est strategy are often olitical andf finstitutiona rather than technical. I they can beovercome, economic savings and improved chan-ces of sustainability maybe achieved.

Ch. 3-Strategies forthe FMure . 75

When and Whe~S&h&ll&~ntrol Programs●

During the recent grasshopper and locustcampaigns vast areas were sprayed with insec-ticides. The high costs of these efforts, includingthe less clearly documented environmental costs,require a reexamination of where and when spray-ing should be done when future outbreaks occur.Some decisions could be worked out ahead oftime, e.g., the level of infestation required forcontrol of the various species, by representativesof African and donor organizations. Alternately,various control strategies could be selected andcoupled with improved plans for carefullymonitoring their impact.

Many experts conclude that early treatment,especially of hopper bands, is most efficient, andthe economic, institutional, and environmentalcosts of control increase with waiting (99). Forexample, carbaryl and malathion are much moreeconomically applied against U.S. rangelandgrasshoppers earl in their life cycle; optimal con-

itrol occurred at t e fourth instar when grasshop-pers were beginning to cause enough crop damage

7to justi control costs yet populations were stillrelative y small so control could be limited (66).

On the other hand, some propose later treat-ment, perhaps waiting until swarms pose an actualthreat to crops and not spraying rangeland andforests at all unless they border threatenedcropland. This approach increases the risk of cropdamage because insects can move quickly and sig-nificant time is required to mount a spray opera-tion. When environmental conditions are right, forexample, gregarious swarms of the Desert Locustappear more or less simultaneously over a largearea (4). Under these conditions insects couldthreaten crops before a spray operation could bemounted. Thus, a late spraying approach may havehigh political costs (71, 121).

Others propose careful review of the lessonslearned in controlling analogous pests, such as theAustralian Plague Locust orquelea birds. Queleabird ovulations can increase rapidly after rains,

Ebut t e control strategy is to kill only those birdsactually attacking crops. Likewise, methodsdeveloped elsewhere to make pest control moreeffective could be applied to locust pro rams. For

[example, general information is availa le on the

relative merits, disadvantages, costs, and uses ofvarious ground- and aerial-spraying methods (95,118). Some

rt surveys have been organized for

international chemical control efforts, but littleinformation is available on nonchemical efforts(37). And few of the recent grasshopper andlocust spray operations were followed by post-ap-plication assessments of numbers of insects killedthat would help in future decision-making regard-ing control tactics.

The U.S. Forest Setice (USFS) developed a

rs tern for monitoring gypsy moth populations toetermine when and whereto mount control and for

assessin control operations to determine which werefmost e ective. This pro am illustrates the type of

fwork needed to improve ocust and grasshopper con-trol. Special “forest pest management” grou s lay

xout plots for gypsy-moth treatment and deci e theappropriate time to do treatment, based on athreshold number of eggpods and stage of develop-ment of the caterpillar. Aerial treatment is doneduring specified weather conditions. Then, the pestmanagement groups revisit anumberof treated plotsat 7, 14, and 21 days to check the number of insectskilled. Usually thesameteamdoes re-and ost-ap-

r“~plicationassessments. Dataonapp lcatlon e.g., for-mulation, characteristics of the e uipment and

l)plane, pilot’s name) and, when possi le, treatmentresults for each plot are remrded on standardizedforms. From this dat~ the USFS learned that resultsdepended significantly on which pilot did the spray-ing, and that treatment should begin at lowerthresholds so that smaller areas could be s rayed

#(59). These methods and lessons may be irectlyapplicable to grasshopper and locust programs.

Resolving issues of when and whereto controllocusts and grasshoppm is USAID’s responsibility.Policymakers need to listen to all sides of thedebate, examine available evidence, and thendetermine ways to be more selective regardingtiming and target sites to reduce costs (includingenvironmental costs) and maximize effectiveness.

WHAT CONTRALTO USE: THE ROLEOF TECHNOLOGY

The choice of technology to control grasshop-pers and locusts, as for other purposes, carries withlt a variety of consequences. Some technologiescan play a strong development role while otherscan hinder development. Often, support for in-


dividual types of technologies sets up complextrade-offs.

The decision to support widespread pesticideuse for agriculture is such a case. In effect, donor-supplied pesticides subsidize high pesticide use.Because of these subsidies, users paid from 85percent to only 10 percent of the real cost ofpesticides in one stud of nine developing

[countries. Users paid on y 11 percent of the realcost in Senegal and 33 percent in Ghana, the twoAfrican countries included; these subsidies wereworth $4 million and $20 million, respectively (80).As a result, farmers have decreased or abandonedalternative control methods–such as soundagronomic practices and varietal selection–infavor of pesticides. The social and environmentalside effects of these chan es are largely undocu-

?mented but may be signi leant. For example, in-creased pesticide use was among the factors thataccompanied the increased commercialization ofagriculture. This process has increased demandson women farmers’ labor, reduced the amount offood grown for local consumption, and en-couraged planting higher value crops.

Today, widespread pesticide spraying is thepredominant technolo used against grasshop-

?pers and locusts. Usua ly, effective pest manage-ment for crops includes a larger number and widervariety of options (table 3-2). Implementing along-term development approach to locust andgrasshopper management requires broadeningthe current range of technolo les and identi~ingor developing ones that can L used by variousgroups in environmentally, economically, and in-stitutionally sustainable ways. Integrated pestmanagement, joined with various forms of earlywarning, are two types of technology that holdpromise. Both require additional research to befully operational.

Integrated Pest Management

Finding: Integrated pest management isUSAID% stutedpolicy, but many elhne~ of such anapproach wem notadkquately emphasized dun”ngtherecent grasshop~r and locust campaigns, partly be-cause of lizck of available technology and partly be-cause of thepoorpeflormance of &nom andAftiagenciks. If USXID tind% b implement &policyfully, the Agency must suppoti mseamh to dkvelbpalternatives to widkspnud spraying, cOh!ect W on

economic injury lkvels ofcmps, assess thee~ectivenessof various control strat@s, and wise its approachbased on these eflorts.

Integrated pest management is “the optimiza-tion of pest control in an economically and ecologi-cally sound manner accomplished by thecoordinated use of multiple tactics to assure stablecrop production and to maintain pest damagebelow economic inju~ level while minimizinghazards to humans, ammals, plants, and the en-vironment. In its broadest form an 1PM pro ram

fencompasses all significant components o theagroecosystem–soil, crops, water, air, insects,pathogens, weeds, nematodes, and other or-ganisms--which interact among themselves andwith other components of the system.” (125).

Integrated pest management combines avariety of control techniques to reduce and keeppest populations at acceptable levels, based oncriteria of crop ield, profit, and safety. It seeks

?’maximum use o biological control, pest-resistantcrop varieties, and cultural practices. Pesticidesare normally used only after the target pestreaches an infestation level called economicthreshold or economic injury level, i.e., a pest den-sity at which the costs of control “ust equal crop

ireturns. Even if insecticides are t e onl control{option available, an 1PM a preach stipu ates that

Fthe chemicals be used as ef ectively and efficientlyas possible and their environmental and healthimpacts be monitored carefully.

Furthermore, 1PM can be described as a wayof thinking, a process of dealing with a problemholistically. This approach requires flexibility andthe ability to deal with multiple factors at one time.Practitioners must be discriminating, adapting thesame principles to different situations, rather thanappl mg a single solution to all cases in a narrow,

zblac or white way of thinking. In this sense,mediating diplomatic solutions to border disputescould be considered part of an 1PM strategy forlocust control in Africa.

Promotion of 1PM is USAID policy. However,it still is not used widely within USAID’s agricul-tural and health projects. The Agency tends tosupport 1PM in special pro”ects rather than in-

Itegrating it into overall deve opment strategy andprograms (22). Many feel that USAID shouldsupport increased research on 1PM and make in-

Table 3-2–Control Tactics Now Employed Against Major Pests of Wheat in the U.S. Great Plains and Sorghum in Texas

7Biolo “cal HostPred.a plant Elimi- Crop

C“l:: :Z::=‘“:,or::::

and Micro- resist- Sanita- nating rota- Planting CleanMajor pests para. bial anee tion hosts tion date

Wheat:Hessian fl b

xGreenbugWheat stem sawfly’Army wormsc

Cutwormsc

Aphidsc

Grasshoppers c

Wheat stem ma otc

FFalse wirewormTrue wirewormcSorghum:White grubWlrewormsGreenbug aphidb

Fall army wormbBeet army wormbS.W. corn borerbSugarcane borerChinch bugSorghum midgebSotwhum webworm

1111111111

1121111111

1111111111

3121111111

2111111111

2221111111

1111111122

2111212111

1111111111

1111111111

1211111111

2121211222

2111112122

1111111111

1313333111

3222212122

1112211111

1111111

“ 11

113111121

221113311

221111121

231112211

222312323

111111111

111111111

111111111

111111111

321111111

131111111

113221123

112111111

111111111

NOTES:~Predators and parasites

introduced pest‘native p3st

KEY: 1 = little or no use2 = some use3 = major use

SOURCE: U.S. Congress, Office of Technology Assessment, Pest Management Strategies in Crop Protection, vol. 1- Summa~ (Springfield,1979, pp!22, 54).

VA National Technical Information Service, October


creased efforts to integrate 1PM in the majority ofits a ricultural programs. Generally, the concept

1?of I M is not well-understood b decisionmakers.i?For example, most USAID of lcials responsible

for the grasshopper and locust program maintainthat 1PM does not apply to grasshopper and locustcontrol during upsurges (44).

However, various elements of 1PM neverthe-less were clearly appropriate during the recentcampaigns and poorly implemented:

Optimization of control-l%is refers to effi-cient and effective use of resources, differing frommaximization of control. The large numbers ofhectares sprayed could have been treated far moreeffectively vwth available technologies. Pin

rint-

ing tar ets, improved consideration of win drift,8groun temperature, time of day, stage of insect

develo~ment–among other things–would havegreatly unproved efficiency.

Multiple control tactics-These were not usedbecause control methods against migrating swarmsare limited. The lack of alternative methods, how-ever, reflects the lack of resources and low prioritygiven to developing them. Donors could have setaside more resources for developing alternativesrather than spending the overwhelming propor-tion of their funds on emergency spraying.

Pest damage kept below the economic injurylevel (EIL) to maintain stable crop production-Major crop loss due to grasshoppers and locustsdid not seem to occur at the national level in 1986to 1989, although some individual farms sufferedsignificant losses (18). By and large, swarms didnot affect croplands. In some cases, sprayingseemed to protect crops. The lack of dama e can-

fnot be attributed automatically to contro, how-ever, because of the complex relationship amen

fincreased rainfall, insect upsurges, and crop yiel .High rainfall in the mid-1980s increased cropgrowth in many areas, making “stable crop produc-tion” difficult to calculate. Reliable data needed tosort out these various factors are lacking so it isalso difficult to determine economic injury levelaccurately. Even so, little, if any, effort was madeto base decisions to spray particular areas on sucha determination.

Minimal hazards to peo le and the environ-/!’ment–Atbest, this element o IPMwas not carried

out consistently, despite efforts by USAID andothers. For example, broad-spectrum insecticideskilled nontarget organisms, and disposal of excesspesticides and their containers remainsproblematic.

Relatively workable 1PM programs have beendeveloped for a range of

rts and cro s and are

r)being used in some deve oping areas 103. Thecost-benefit analyses of those pro~rams evaluatedgenerally show a reduction in pesticide use and anincrease in profits (35). 1PM has not been em-phasized in locust and grasshopper control inAfrica and the Middle East, however (95). Today,biological control, cultural practices, and othernonchemical components of 1PM cannot providethe high level of control needed to stop gregarioushopper bands and swarms of adults. Thesemethods might, however, contribute significantlywhen used together or at early stages of an infes-tation (9).

An effective 1PM program would aim to

frevent serious locust and grasshopper outbreaks.t could include activities at a variety of levels, but

regional aspects would be necessary due to thecross-boundary migration of insects. New 1PM ap-proaches would rely on controlling locusts andgrasshoppers at earher points than achieved in therecent campaign, similar to the “strategic control”advocated by FAO for the Desert Locust, butplace a greater emphasis on using alternatives tospraying as these become known or available.

Examples of 1PM strategies for grasshoppmand locusts might include planting alternative cropsthat are less susceptible to these meets; increasinganimal production; developing cottage industries to

Freduce locust meal for food or to produce extractsrom neem trees for use as an antifeedant (126 , and

2developing pesticide regulations to improve c emi-cal use. Sound land management-especially refores-tation, upgrading range quality, and avoidance ofovergrazing and widespread burning-can suppress

%rasshoppers and locusts and decrease suitable

$reedingsites 95). This andotherapproaches might

be part of an I M approach for some other speciesas well.

Ch. 3-Strate@”esfor the Futire ● 79

Certain as ects of an 1PM a roach tox Pgrasshopper an locust problems cm be imple-

mented immediately, e.g., improved use of pesticides.In the short-term, improved regulation, selectio~storage, application, and disposal of pesticides maybethe best strategy, especially for reasserting controlafter an u urge (95). Mechanical and cultural

rmethods o control are also currently available andthese might be suitable for controlling small infksta-tionsincro .Theyaremostlikely tobeusefulforthe

2Variegat Grasshopper, especially ifpairedwith ad-ditional training for extension agents.

Research on microbial and botanical pes-ticides, insect ~pulation modeling, forecasting,developing resistant crop varieties, and furtherimprovements in insecticide application offer abetter outlook in the medium and long-term 95).

$Distinct approaches will have to be develope foreach of the major locust and grasshopper species,however. For example, since the Desert Locusteats many types of vegetation, developing resistantplant varieties does not seem to be a feasible ap-proach to controlling it.

Biological Clmtml

Normally, naturally occurring biological con-trol is not sufficient to prevent outbreaks of majorlocust and grasshopper species (93). But enhancedbiological control-the use or encouragement ofnatural enemies for the reduction of pests-is onepotential component of an improved 1PM ap-proach. Ikcusts and rasshoppers have anarrayof

fnatural enemies. So ar, these have not been usedin control campaigns, nor has what is known aboutnatural pest mortality been ex loited to produce

rpredictable or consistent resu ts (95). Some feelthat biological control offers considerable poten-tial, although additional research and field testingare required before their real value will be known.Because of the priority currently given to chemicalcontrol, much of the research on alternativemethods is in its early stages.

Some biological control agents, when pack-aged, are called microbial pesticides. Most havethe advantage of easy deployment; they could beformulated and sprayed or used as baits in muchthe same way that chemical insecticides are now.Some newer biotechnology may be helpful indevelo ingthese alternatives. However, microbial

rcontro s require EPA registration for commer-

cialization and such approval is difficult to obtainfor emetically engineered microorganisms.

fSimi arly, African governments want reassurancethat these biological control agents do not posehazards to human or animal health.

Grassho pers and locusts are susceptible toEinfection by acteria, viruses, fungi, and protozoa

and several potential new microbial controlmethods are being tested. Nosema Zocustae, thefirst protozoa registered by EPA for use against aninsect, is approved for control of U.S. rangelandgrasshoppers. Developed at USDA’s AgriculturalResearch Service’s Ran e Insect Control Re-search Unit in Bozeman, hontana, it is sold com-mercially as Nolobait. Used with a wheat-bran bait,it takes 3 to 4 weeks to kill 50 to 60 percent of theinsects and persists for two seasons because it ispassed from one generation to another. It is lessexpensive than chemical insecticides and does notadversely affect beneficial species or other naturalenemies (21, 88). Field experiments in Cape Verdeand Mauritania showed that native grasshopperswere infected with Nosema (39) but did not deter-mine whether it could suppress grasshopper out-breaks (9). USAID supported Nosema research inMali; it was stopped in 1988 due to MalianGovernment fears of possible hazards (99).USAID supports further work on Nosema andother microorganisms in Cape Verde by USDAscientists and the national agricultural researchservice. Several recent studies suggest that furtherresearch in Africa on various s ecies of Nosema

rmay a off for rasshopperand ocustcontrol (95,99). fi?DA an~ other researchers began examin-ing viruses as potential control agents becauseviruses are more deadly, kill faster, and could beused in combination with slower-acting microbial.For example, an entomopoxvirus for the Senegalesegrasshopper shows potential as a microbial controlagent (94). The fungal pathogen Entomophaga gryl-li attach some locusts and grassho pers. It has not

Cfbeenstudied in Africa orthe Mid leEast (95), butitspotentia.1 insemi-aridareaswhere mostgrasshoppmoccur seems small because fungal developmentde nds on high humidity (94 . It may be usefid inx )‘ca’s humid areas, however, or these same reasons.Some new strains of spore- or toxin-forrnin bacteria

f(like those used already for biological ccmtro for other.msects)mightbeisolatedfromlocustsandgrasshop rs

r(78). Rickettsia are virulent to grasshoppers, but t eiruse may be too hazardous to have much potentialbecause they also infect vertebrates (94).

80 ● A Plague oflzxusts

Other Biorational Contmik MatzriuLs

These include botanical pesticides andpheromone traps-alternatives to Synthetic chemicalinsecticides. One botanical insecticide has receivedattention, especially for its antifdant effects. Ex-tracts from neem trees (Azadirachta indica) dis-coura e loctits,

! Yasshoppers, and other insects

horn eedingon pants to which it is applied (9 . In2India, neem spray and dust protected crops om

Desert Locusts an~ in Togo, neem repelled ~as-shoppers. However, 1988 trials at InternationalCrops Research Institutefor the Semi-Arid Tropics(ICRISA in Niger were less than successful and

2indicated t at farmers might be unwilling to investthe labor or funds to use neem orJ grain crops, sincerepeat ap lications are needed (99). A neem insec-ticide, Jargosan-0, is being distributed in theUnited States by W.R. Grace and G., but EPA hasnot ap roved its use for food crops. The authors of

bUSN ‘S Pro rammatic Environmental Review2and the AEL A evaluation supported further re-

search on neem as an antifeedant.

The E tian Government supports researchTon the anti eedant properties of a number of in-

digenous plants, and the German Agency forTechnical Cooperation (GTZ) funds trials withneem, Nosema, and other natural a$ents as part ofits program of developing altematwe methods oflocust and grasshopper control (107).

The International Center on Insect Physiologyand Ecology (ICIPE) and others are attempting toidentify natural attractants. Recently, ICIPEachieved some success using pheromones (natural

kattractants as bait to trap certain species of thetsetse fly ( ashington Post, April 3, 1989). Likebiological control agents, attractants are usuallynarrow-spectrum and thus less harmful to nontar-get organisms and the environment than broad-spectrum chemical insecticides. The potential forusing pheromones for grasshop r or locust con-

rtrol 1s not known and many fee that pheromonework is not justified for this reason (6).

Nw Reseamh on Alte~”ve Cbntmk

Those en aged in planning and conducting re-fsearch on bio ogical control agents, especially the

microbial ones, stress that it maybe 8 to 10 years or

longer before these will be ready for large-scaleuse (55, 65). First, the microorganisms have to beidentified and isolated from locusts and grasshop IX

rtih(~).~enwriowfo~~atio~ m~t~ eldtested against target s~ies and nontarget or anisms

funder various conditions and these resu ts cor-roborated. Fmall , wa~ to mass-produce and applythe agents must L developed and tested. Researchprojects such as these reqture long-term institutionalsupport for an agency to attract qualified scientists andsustain their wrk.

The International Institute for Tropical Agricul-ture (IITA) recently be an a major research effort

fon biological control o grasshopper and locusts.The $1.0 million USAID-fimded project aims todevelop strains of two fungal ~atho~ens recoveredfiomlocustsand assho persmAfnca as biological

Er{ticidesandfie dtestt eminthe Sahel. WorkWill

led by scientists from the London-based Com-monwealth Agricultural Bureau International’s In-stitute for Biological Control at IITA’s facility inBenin.

1cpEahzrOra major research initiative.

By late 1989, I IP had received $0.5 million hornthe World Bank and African Development Banktoward the $14 million r uested for the first 5-year

%hase, 1989 to 1993. ICIP ‘s proposal encompassesEve areas of research on alternative control metho@including biorational agents and improved chemicalinsecticides:

population dynami~ (to detect potentiallydangerous populations during recessions);

pheromones and kairomones (to use asattractants in locust control);

endocrinology of locust phase<hanges andgregarious behavior (to pinpoint targets forgrowth regulators and broad-spectrum chemical. .~)

.9

biological control (to augment role ofpathogens and arasites, includin enhancin

! ftheirvu-ulence ygeneticmanipu ation); anf

new approaches to the use of baits (since theytend not to affect natural enemies andnontarget organisms).

Ch. 3-Stratep”esfor the Future ● 81

Monitoring Insects, Weather, and Vegetation

Finding: Technologies for ground monitoringinsect popuhtims are adequute but sometimes amused ine+i’vely. Techno&@s for monitoringfimthe air tend to be impmcise and their msulks ofindelivered lati?. Themfom, technological and institu-twnulimprovementsw needkdforgmundandaerialsurveillance andfomcdng, necessary wmponenO ofa preveti”ve strategy.

Monitoring is essential for a number of purposes.A preventive ap roach to locust and grasshopper

?control requires orecasting, ground monitoring, andearly treatment to interrupt swarm formation. Effec-tive pest management strategies require monitoring,or tracking insect ~pulations before control to findident~, and delinut infestations and further monitor-ing after control to assess its effectiveness. Famineearly warning systems benefit horn information onfluctuating insect populations.

Ttzhnolbgiks

Methods already exist for monitoring pestpopulations on the ground and for measuring theImpacts of control but their use needs to be im-proved, especially by increasing national capacity.

Today, most remote sensing and forecastingwork is done by expatriates at scientific centers inEurope, the United States, or regional centerswithout ad uate, timely, and accurate field data.

7Consequent y, African field programs remainlargely untouched by the technological advancesat remote sensing centers; quickly exchangin in-

!formation between the field and centers is dif lcult(95); and often forecasts are wrong.

Anarrayof detection strategies, each appropriateforse

“ c times and locations, can im rove forecast-ing. Jmeinforrnationcanbeobtain only bygroundsurv teams (insect species, stage of development

7popu ation density). Other information can be ob-tained best from amxaft and satellites (current andlikely future vegetation, wind and rainfall patterns).Combining remote sensing data with maps showing:1) political boundaries, roads, and landmarks, 2)historic breeding areas and migration patterns, and3) insects’soil andvegetationpreferences can be usedto help ~ound survey teams select high priority areasformomtoring. (George Popovpreparedmaps on thepreferred habitats of the Desert Imcust in the Sahel

for FAO but these are not yet available to nationalcrop protection services.)

All aerial survey methods require roundfverification. Thus, they cannot substitute or cru-

cial ground monitoring and improved integrationof the two methods is critical. For example, infor-mation from remote sensing could better guide thework of ground teams just as insect populationdata from ground teams could supplement thevegetative cover data provided by remote sensing.

The most critical component of early detec-tion of pest populations is a network of trainedground observers (37) with adequate equipment.Thus, training remains one of the most importantneeds for improved field applications of forecast-ing. Training could encourage managers to makegreater use of remote sensing and provide a cadreof field officers for various early warning and sur-vey activities, including data interpretation (95).Certain aspects of monitoring programs are un-resolved. For example, some feel that a monitor-ing system designed for pest complexes would bea more efficient use of resources than onesdesigned for single insect pests. Any effective sys-tem, however, must include many levels of or-ganizations, working within the framework ofnational and regional programs, to improve ac-curacy and sustainability.

Types of Early W~ing and Forecasting Systems

Current early warning systems combine remotesensing data with other aerial, ground and statisticalinformation for a variety of purposes, such as agricul-tural and environmental assessment and resource

Tdata, for example,management (45). AGRHYMEare used for crop and pasture monitoring in theSahel.

Several groups monitor pest damage as one ofseveral major risks to agricultural production topredict food shortages and famine, and thus an-ticipate the need for food aid and other forms ofassistance. USAID’S Famine Early Warning Sys-tem (FEWS) and FAO’S Global Information andEarly Warning System are examples.

Three major organizations make or plan to makelocust and grasshopper forecasts s

rifically: 1 )

FAO/ECLO tkXl@ t h e ARTEMI (~ca Real-Tii Environmental Modeling Using Imaging Satel-


lites) proj~ 2) the French research age~, p~~(pf?ro arnmede Recherches Interdisciplinary Francaissur es Acridiens au Sahel, reorganized now asAcridologie Operationnelle-Ecoforce Intemationale),and 3) the Permanent Interstate ~mmittee forDrought Control in the Sahel (CILSS) meteorolo~agency, AGRHYMET (99). ‘Ihese type of programshave significant potential. For exam Ie, a model

fpredicting upsurges and locations o the AfricanMigratory Lncus~ developed by a pint FAO/U.N.Development Programme proje@ reduced annualscouting efforts from 144 to 90 pemon-months (2).

Current programs also have serious limita-tions. Reports from PRIFAS and ELCO often arenot quantified, detailed, or timely enough to beuseful in the field. For example, Operation SAS(surveillance des Acridiens au Sahel) was estab-hshed within the French PRIFAS for rapid colle-ction of field observations from a Sahel-widenetwork. However, data collection has been slow,sporadic, and incomplete, preventin reliable

J8prediction (99 . Also, the biweekly SA newslet-

ter has been istributed too slowly for recipientsto use it for planning; it is used primarily as asituation summary. SAS first constructed a predic-tive model for the Senegalese Grasshopper andused historical records, G. Popov’s qualitativevegetation and soil maps, and AGRHYMETweather data (often relying on 30-year averages)but not remote sensing data. In the past 3 ears,

JPRIFAS has been developing a similar m el forthe Desert Locust and is working withAGRHYMET to set u a locust survey and warn-

Aing service for the CI S countries (75).

The ECLO in FAO/Rome provides faster in-formation because its monthly “Desert LocustSummary” is sent by fax. FAO combines data fromfield reports and remote sensing. Originally, FAOused Landsat data, but now uses Meteosat andNational Oceanic and Atmospheric Administration(NOAA) imagery in the Dutch-designedARTEMIS system. FAO also uses this technologyto produce 10- and 30-day rainfall maps, relying onthe European Centre for Medium-Range WeatherForecasting for forecasts of temperature, pressure,wind, and rain for up to 5 days in advance (13). Likethe SAS Bulletin, however, FAO’S “Desert Imcust

~~~~~~eld&ta(95).“ is lagued by gaps in coverage due to

FAO’S separate “Uc

ate” includes a generalstatus report, a l-mont forecas~ descriptions ofweather and ecological conditions, specific countryinformation on pests sighted and assistance re-ques@ and assistance provided bydonors. Recent-ly, ECLO entered historical data on locust plaguesin its computerized database and plans to use it inforecasting locust migration patterns.

Retnoti Sensing and Greenness Mizps

Satellite-based weather, vegetation and landSUrVw, maps, etc., are all likely to be useful forbuilding scientific institutional capacity in Africancountries. Such information can be used for gover-nment @arming and regulation and for monitoringdesetitication, vegetation, surface features, windpatterns, etc. Probably satellite-based remote sens-ing will be used less for locust and grasshopperforecasting and control than for these purposes. In1988, the multidonor Club du Sahel commissioned astudy of 50 remote sensing projects in the Sahel.Remote sensing seemedve useful forclimatologi-

7cal applications, less useful or crop monitoring (al-though ve etation indexes were of some use), and

fleast usefu for forecastingyields because ofdifficul-ties in measuring crop acreage and discriminatingbetween crops (67).

USAID sponsored the development of green-ness maps, one particular type ofve etation index,

5by the U.S. Geological Survey (U GS) in 1987.Greenness maps were furnished to five Saheliancountries every 2 weeks between 1987 and 1989 bythe USGS EROS (Earth Resources ObservationSystems data center in South Dakota, using data

Jfrom N AAsatellites. These maps showed chan-gesinvegetation overtime. FAO’s ARTEMISpro-gram also monitors rainfall and changes inve~etativecover. These maps helped field teams ldentifjplaces where locusts might be found and areaswhere ground surveillance was not needed (95),especially in places where rainfall is irregular andground cover inconsistent.

The USGS greenness maps were valued highlyby those interviewed during the AELGA evalua-tion but were judged not too useful for makingcontrol decisions because delivery to Africa tookup to 2 weeks (in 1987) or 8 days (in 1988). As aresult, maps were sent by fax to Mauritania and

Ch. 3-Strategies forthe Futim ● 83

Niger by late 1989 (121). Both USGS and theARTEMIS maps have another weakness that isless easily corrected. Areas with very low amountsof vegetative cover may not show up on existingsatelhte imagery yet be areas where potentiallydamaging Desert Locust populations develop(13).

Imagery for grasshopper and locust control isor can be provided by several types of satellites:

● Meteosat, operated by the European SpaceAgency;

. weather satellites operated by NOAA (partof the U.S. Chnmerce Department);

. Landsat, developed by the National&ronautics and Space Ahninistration butowned since 1984 by the private U.S. EarthObservation Satellite Co.; and

. the French S teme Probatoire d’ObsemationTdelaTerre( POT) (figure 3-1).

The first two are used by those monitoringinsects now; the second two provide more detailedinformation on land cover. Landsat has greaterresolution than NOAA’s polar orbitin satellites

fbut NOAA provides daily coverage whi e Landsatpasses over the same areas only once every 16days.Landsat has not proven capable of monitoringcrop production (26) and obtaining Landsat datais more expensive than from NOAA satellites soFEWS and USGS rely on NOAA’s system. Ingeneral, a confusing array of Earth-monitoringsatellites exist, and the U.S. Government has beencriticized by scientists and others for having spenttoo much on satellite hardware that produces toomuch inaccessible and unanalyzed data (56).

USAID plans to transfer significant aspects ofU.S. remote sensing application to locust forecast-ing to African countries or regional organizations(62). USGS, whichhassup rtedAGRHYMETfor

ra number of years, recent y trained AGRHYMET

staff and key personnel of the Sahelian nationalcrop protection services to use greennas maps.Also, USGS technic ians are t ra in ingAGRHYMET staff to produce and distributetheir own greenness maps (99). AGRYHMET isexpected to provide this service to its nine memberstates in 1990, according to some sources (45,62 ,

)or within the next3 years, according to others (99 .Similarly, USGS is transferring greemess map-making capability to Tunisia for Northwest Africanand planning to develop it in Djibouti for the sixEast African nations (62). USAID is funding in-stallation of a satellite dish in Niger soAGRHYMETwill be able to receive data directlyfrom the NOAA weather satellites.

Currently, remote sensing for early warning of

Frasshopper and locust upsurges is not consideredully operational nor does rapid transmission from

satellite to Earth ensure that all sta es of datafgatherin , analysis, and use are coor inated and

frapid (9 ). One perceived danger is that, as theseprograms develop, remote sensing will dominateother types of information-gathering, thereby reduc-ing the resources available for field scouting. Forexample, obscxvers are concerned that FAO’S inter-est in a very expensive, centralized program based inRome ma preclude other, less glamorous, ap-proaches. 6n the more promising side, plans exist toextend satellite-based monitoring to other impor-tant migratory pests such as the grain-eating queleabird, the African Migratory bust, the SenegaleseGrasshopper, armyworms, and the Red Locust (95).

The various groups conducting early warningand remote sensing activities do not necessarily

i?duplicate efforts because they operate with di -ferent mandates for research, applications, infor-mation dissemination, and training. Nevertheless,clear duplication of effort exists and improvedcoordination and cooperation is needed (95). In-ternational organizations are mo6t suited to providesup~rt for remote sensing, due to the high mst ofeq.upment and thecomplexityofsupport semices, butregional groups might be responslkie for establishinguniform reporting systems.


Figure 3-1-Principal Satellites Used in Early Warning and ForecastingMeteosat (geosynchronous)35,800 km. altitudeImages hemisphere every30 minutes

10altitude

e

w

SPOT832 km. altitude26 day repeat cycle(more frequent imaging ofselected areas upon comm

.:,f K~~thwidthB’85km ySwath width 2700 kilometers ~ “.”j:;:..’. . “, ,“. . . . . . . . . . .. . , .. . . . : “. . . . ., . . . , ,,.“,,,. .. . . . . . . . . . . . . . . . . . . . - . . . . “. “. . .“ .“ “ . . . . .. . , . ,. ., . . . . . ., . . . . . . . . . . . ..- . . - . “. . . . . . ., . .. . . . . . , . . . . .’ ..-

. . ‘ . . . ,..-Image Resolution

10 meters, 0.01 ha. panchromatic bands SPOT20 meters, 0.04 ha. multispectral bands SPOT

‘30 meters, 0.1 ha. thematic mapper (TM) Landsat*—80 meters, 0.5 ha. multispectral scanner (MSS) Landsat

Advanced very high resolution radiometer (AVHRR)+ NOAA 9 and 10 satellites >

1 1,1 kilometers, 120 ha. local area coverage (LAC)4.0 kilometers, 400 ha. global area coverage (GAC) 1

SOURCE: TAMS Consultants, Inc. and Consortium for International Crop Protection, Locu.wand Grasshopper ControZinAj7ica/Asia: AProgramrna tic EnvtionmentalAssessmen4 Main Report, contractor report prepared for the Ageney for International Development,March 1989, p. D-7.

Documents

Strategies for the Future - Princeton