Transactions of the Royal Society of Tropical Medicine and Hygiene (2009) 103, 1263—1268

avai lab le at www.sc iencedi rec t .com

journa l homepage: ht tp : / /www.e lsev ier .com/ locate / t rs tmh

Relationship between coffee cultivation practices inColombia and exposure to infection with Leishmania

Bruce Alexandera,∗, Luz Adriana Agudelob, Jose Fernando Navarrob,Jhon Fredy Ruizb, Jorge Molinac, German Aguilerac,Adriana Kleind, Martha Lucia Quinonesb,1

a Instituto de Ciências Biológicas, Universidade Federal de Minas Gerais, Caixa Postal 486,Belo Horizonte 31270-901, Brazilb Programa de Estudio y Control de Enfermedades Tropicales (PECET), Facultad de Medicina,Universidad de Antioquia, Medellin, Colombiac Centro de Investigaciones Microbiologicas y Parasitologicas (CIMPAT), Universidad de los Andes, Bogotá, Colombiad Universidad Industrial de Santander, Ciudad Universitaria, Carrera 27 Calle 9,AA 678, Bucaramanga, Santander, Colombia

Received 28 July 2005; received in revised form 24 April 2009; accepted 24 April 2009Available online 24 June 2009

KEYWORDSCutaneousleishmaniasis;Leishmania;Incidence;Transmission;Coffee;Colombia

Summary The inhabitants of coffee-growing municipalities consistently report the highestannual rates of cutaneous leishmaniasis in Colombia. During the last two decades most Colom-bian coffee growers have changed from the traditional system of cultivation, where the cropis grown under different species of shade trees, to an intensified system where it is grownat high densities in full sunlight. This change may affect transmission of Leishmania spp. tohumans in several ways, probably resulting from reduced human—vector contact. The responsesof residents of traditional and intensified coffee plantations to the leishmanin skin test werecompared to ascertain whether intensification has indeed affected Leishmania transmission.Although prevalence of infection was significantly higher (P ≤ 0.01) among residents of tradi-tional plantations (26.8%) than among those of intensified ones (13.2%), no significant differencecould be demonstrated with respect to incidence of infection at the time of the study. Similar

rates of infection were found for men and women, although the incidence of infection wassignificantly higher among the latter in intensified plantations. Changes to the type of datacollected and the data collection process will facilitate the evaluation of the long-term effectsof intensification of coffee plan© 2009 Royal Society of Tropicareserved.

∗ Corresponding author. Present address: Liverpool School of Tropical ME-mail address: bruce.alexander@xeroshield.co.uk (B. Alexander).

1 Present address: Depto. Salud Pública, Facultad de Medicina, Univers

0035-9203/$ — see front matter © 2009 Royal Society of Tropical Medicindoi:10.1016/j.trstmh.2009.04.018

tations on Leishmania transmission.

l Medicine and Hygiene. Published by Elsevier Ltd. All rights

edicine, Pembroke Place, Liverpool L3 5QA, UK.

idad Nacional de Colombia, Bogotá, Colombia.

e and Hygiene. Published by Elsevier Ltd. All rights reserved.

1

1

CttCptttteeabotmnpa1

tCbtfhtotsiowmlftavrrwhmbtrCbfi

LlItatsd

oren

motstr(oo

2

2

TdPi((wCoCEmcstmoopiosatsz

2

Tataalta

264

. Introduction

olombia is the world’s second largest coffee producer,he 558 000 tons harvested in 1998 (10.1% of the worldotal) being exceeded only by Brazil. Coffee provides 37% ofolombia’s agricultural employment, involving 3.3 millioneople on 566 000 farms in 590 municipalities.1 Accordingo the 1997 coffee census, a total of 893 445 ha (0.8% ofhe total area of Colombia) was under cultivation in 16 ofhe country’s 32 departments. The importance of coffeeo the Colombian economy has steadily fallen from 54% ofxports in 1954 to only 12% in 1998, but as a source of exportarnings it is still exceeded only by petroleum. Althoughutochthonous cases of cutaneous leishmaniasis (CL) haveeen recorded from all but three of the 32 departmentsf Colombia, at least 1700 (29.7%) of the 5716 reportedo the Ministry of Health in 1994 were from coffee-growingunicipalities. Ten of the 20 municipalities with the highest

umber of reported cases of leishmaniasis were in coffee-roducing areas of the department of Norte de Santandernd 8375 cases were recorded from this department in990—1997.

The arrival in 1984 of coffee rust fungus Hemileia vas-atrix was the principal motive for the intensification ofolombian plantations, a process by which bushes of the ara-ica bourbon and tipica varieties grown in the shade of largerees (traditional cultivation) were replaced by those of theungus-resistant colombia hybrid strain, grown intensively asigh density monocultures in full sunlight (intensive cultiva-ion). Shade plantations often represent the only remnantsf the original forest in a particular area. They are rela-ively small (1.61 ha being the mean area in Colombia) butupport a rich diversity of plant and animal species, includ-ng phlebotomine sand flies (Diptera: Psychodidae), severalf which transmit Leishmania spp. to humans.2,3 Since theild mammals bitten by sand flies include potential Leish-ania reservoirs such as rats and opossums4 and growers

ive adjacent to their coffee crops, all the factors necessaryor transmission of the parasite to humans may be concen-rated within a small area. Intensified plantations wouldppear to represent a far less hospitable environment forectors and reservoirs of Leishmania. Removal of shade treeseduces the production of leaf litter and affects soil fertility,equiring greater use of chemical fertilisers. Weeds whichould be unable to grow in shade have to be controlled byerbicides and as monocultures intensified plantations areore accessible to insect pests. The problems of the Colom-ian coffee industry were exacerbated with the arrival ofhe coffee berry borer (Hypothenema hampei) in 1988. Thisepresented the first serious insect pest of coffee to reacholombia and has become so widespread that growers haveeen forced to use insecticides such as endosulphan for therst time.5

A possible relationship between coffee cultivation andeishmania transmission was first noted by Scorza and col-eagues in Venezuela6,7 and Warburg et al.8 in Colombia.n both countries optimum conditions for coffee cultiva-

ion occur in mountainous areas between 1000 and 2100 msl, where cutaneous leishmaniases caused by members ofhe Leishmania (Viannia) braziliensis and Le. (Le.) mexicanapecies complexes are the most important vector-borneiseases.9,10 Alexander et al.11 also found a high prevalence

ar≥o

B. Alexander et al.

f leishmanin skin test (LST)-positive individuals amongesidents of coffee plantations in southeast Brazil, wherenvironmental conditions are very different from those inorthern South America.

The relationship between coffee growing and the trans-ission of Leishmania spp. has thus undergone a number

f changes during the past 20 years, most of them likelyo be detrimental to the parasites’ sand fly vectors andmall mammal reservoirs. The possibility therefore existshat intensification of coffee plantations in Colombia couldeduce the incidence of American cutaneous leishmaniasisACL) as a collateral benefit. We investigated whether ratesf Leishmania infection differed between human residentsf the two plantation types.

. Materials and methods

.1. Study areas

he study was carried out in seven municipalities in twoepartments of Colombia: Arma/Aguadas (5◦36′N; 75◦32′W),ensilvania (5◦23′N; 75◦9′W) and Salamina (5◦24′N; 75◦29′W)n Caldas; and Bochalema (7◦36′N; 72◦39′W), Gramalote7◦53′N; 72◦48′W), Toledo (7◦18′N; 72◦29′W) and Labateca7◦18′N; 72◦30′W) in Norte de Santander. The municipalitiesithin each region are <50 km apart. The departments ofaldas, Quindio and Risaralda lie on the Central Cordilleraf the Andes and together form the region known as the Ejeafetero or Coffee Axis, while Norte de Santander lies on theastern Cordillera bordering Venezuela. These two depart-ents were selected because of their socioeconomic and

ultural differences, to ensure that the study was as repre-entative of the overall situation with respect to Leishmaniaransmission in Colombian coffee plantations as possible. Nounicipalities from the three other coffee-growing regions

f Colombia were visited. Bochalema was visited on onlyne occasion (December 1998) as the number of intensifiedlantations there was found to be negligible. The remain-ng six municipalities each contained substantial numbersf both types of plantation and were visited on two occa-ions between January 1997 and November 1999, with visitst least 6 months apart. All coffee plantations included inhe study were between 1170 (Gramalote) and 1750 (Pen-ilvania) m asl, i.e. within the lower marginal and optimalones for coffee growing in the Colombian Andes.

.2. Leishmanin skin testing

he prevalence and incidence of Leishmania infectionmong residents of coffee plantations was based on applica-ion of the LST.12 In total, 1145 plantations (875 traditionalnd 270 intensified) were visited and 100 �l of leishmaninntigen injected intradermally into the underside of theeft forearm of each of the residents who consented to par-icipate in the study or gave permission for the test to bepplied to their children.13

The prevalence of Leishmania infection was calculateds the percentage of individuals giving a positive skin testeaction (LST+ve, manifested as a swelling with diameter5 mm 48—72 h after application) among residents of eachf the plantation types. The size of the indurations was

lomb

iwraAtiPrddutNwocoBL

hSmtc

3r

NfcLti

Coffee-growing practices and Leishmania transmission in Co

measured using the ‘ballpoint pen technique’.12 The linesdrawn to delimit the margins of the indurations were cov-ered with a piece of transparent adhesive tape which wasthen torn off and stuck to the data collection form, allowingthe impression to be used as a permanent record.

The incidence of Leishmania infection was obtained fromthe number of positive skin test reactions during the secondsurvey, calculated per 1000 inhabitants per year. In total,1533 residents who had given a negative reaction during thefirst survey were tested a second time, 1032 from traditionalplantations and 521 from intensified ones.

The percentage of individuals consenting to participate inthe first survey was very high, varying from 96.0% in Aguadasto 98.3% in Toledo. Almost all of the people encounteredduring the second survey agreed to be tested again, withlevels of acquiescence ranging from 98.9% in Salamina to100% in Toledo.

2.3. Statistical analyses

The total numbers of people tested and LST+ve individualsin each plantation type and municipality were compared by�2 analysis, as were positive individuals of each sex. Pooledresults were compared by Mantel-Haenszel �2 analysis. Indi-viduals giving LST+ve results in the first survey who hadimmigrated from other parts of Colombia were not includedin the analyses. Statistical comparisons of mean ages of res-idents and LST+ve individuals were made using Student’st-test. These were carried out using the Statgraphics 3.1programme Statgraphics Technologies Inc., Warrenton, VA,USA, which involved a test of normality before the analyses.

3. Results

3.1. Prevalence of Leishmania infections between

residents of traditional and intensified plantations

A total of 2599 residents of traditional plantations (including439 in Bochalema) was skin tested during the first surveyvs. 838 on intensified farms. The prevalence of Leishmania

tTfp3

Table 1 Prevalence of Leishmania infection among residents of C

Department Municipality No. tested

TR IN

Caldas Pensilvania 125 273Aguadas 381 42Salamina 271 273Total 777 588

Norte de Santander Bochalema 439 3Gramalote 489 88Labateca 369 115Toledo 525 44Total 1822 250

Overall total 2599 838

LST: leishmanin skin test; TR: traditional plantation (shaded); IN: intena Assessed by LST (the survey included individuals with antecedents of

ia 1265

nfection in each municipality visited (including Bochalema,hich was only visited once) is shown in Table 1. When the

esults from the six municipalities were pooled, there wassignificant difference between the two plantation types.

lthough the proportion of LST+ve individuals in each ofhe municipalities was higher in traditional plantations thann intensified ones, this difference was only significant inensilvania and Toledo. Note that these two municipalitiesepresent the two geographically distinct coffee-growingepartments in the study. When the results of the twoepartments were pooled separately and prevalence val-es between the two plantation types were compared,he differences remained significant for both Caldas andorte de Santander. The highest prevalence recordedas in traditional plantations of Labateca and the lowestn intensified plantations in Salamina. Prevalence waslose to or greater than 30% on traditional plantationsf five municipalities (Pensilvania, Gramalote, Toledo,ochalema and Labateca), as well as intensified farms inabateca.

The overall prevalence of Leishmania infection was muchigher on the coffee plantations of both types in Norte deantander than Caldas. The difference in prevalence wasore marked in the latter however, that of traditional plan-

ations in the Caldas being 2.0 times that of intensified ones,ompared to only 1.2 in Norte de Santander.

.2. Incidence of Leishmania infections betweenesidents of traditional and intensified plantations

o significant differences could be demonstrated at P ≤ 0.05or any of the six municipalities (Table 2) with regard to sero-onversion between the two visits. Seroconversion rates (no.ST+ve individuals/1000 persons/year) were higher amongraditional plantation residents in three of the six munic-palities, Pensilvania, Aguadas and Toledo, approximately

he same in Gramalote and lower in Salamina and Labateca.he seroconversion rate for traditional plantations variedrom 82 (Gramalote) to 238 (Salamina) per 1000 personser year, while that for intensified plantations varied from9 (Toledo) to 285 (Salamina). The overall seroconversion

olombian coffee plantations at baseline surveya

% LST+ve TR vs. IN

TR IN �2 P-value

32.0 12.1 21.39 <0.00111.5 9.5 0.02 NS7.7 5.9 0.50 NS

13.5 9.0 M-H �2 15.65 <0.00133.5 0 — —28.0 21.6 0.70 NS35.8 29.6 1.24 NS33.3 11.4 8.07 <0.0132.5 23.2 M-H �2 7.08 <0.0126.8 13.2 M-H �2 22.99 <0.01

sified plantation (unshaded); NS: not significant.cutaneous leishmaniasis).

1266 B. Alexander et al.

Table 2 Incidence of Leishmania infection among residents of Colombian coffee plantations at follow-up surveya

Department Municipality No. tested No. LST+ve/1000/year TR vs. IN Relative risk

TR IN TR IN �2

Caldas Pensilvania 53 175 142 91 0.10 1.47 (0.47—4.57)Aguadas 207 28 133 82 0.02 1.76 (0.24—12.93)Salamina 173 205 238 285 0.21 0.85 (0.49—1.45)Total 433 408 85 93 M-H �2 0 0.92 (0.61—1.57)

Norte de Santander Gramalote 246 39 82 81 0.11 0.95 (0.22—4.09)Labateca 163 47 109 189 0.76 3.13 (0.44—22.26)Toledo 190 27 83 39 0.83 0.58 (0.23—1.45)Total 599 113 77 80 M-H �2 0.03 0.96 (0.48—1.98)

Overall total 1032 521 80 90 M-H �2 0 0.98 (0.62—1.54)

LST: leishmanin skin test; TR: traditional plantation (shaded); IN: intensified plantation (unshaded).a Assessed by second LST at follow-up.

Table 3 Prevalence and incidence of Leishmania infection [based on leishmanin skin test (LST)] by age among residents oftraditional and intensified coffee plantations in seven Colombian municipalities

Plantation type T P-value

Traditional Intensified

Mean age of residents at t = 0 28.9 21.7 4.13 <0.001Mean age of LST+ve individuals at t = 0 36.5 32.6 2.24 0.025

74

.

rr

3iis

Dfpprfiot

w(

psttsoct

Mean age of residents at t ≥ 6 months 26.Mean age of LST+ve individuals at t ≥ 6 months 36.

t = 0: first survey; t ≥ 6 months: second survey; NS: not significant

ate was approximately the same in both regions and whenesults from all six municipalities were pooled.

.3. Prevalence and incidence of Leishmanianfection among residents of traditional andntensified coffee plantations according to age andex

ata on LST+ve prevalence and rates of seroconversionor participants in the two surveys, by age and sex, areresented in Tables 3 and 4. The overall mean age of

articipants in the study was significantly higher amongesidents of traditional plantations than those of intensi-ed ones. People with a LST+ve response were significantlylder (P ≤ 0.01) than the general population in both planta-ion types. The LST+ve residents of traditional plantations

otAlf

Table 4 Prevalence and incidence of Leishmania infection [bastraditional and intensified coffee plantations in seven Colombian m

Pl

Tr

M

% LST+ve individuals at t = 0 30No. LST+ve individuals−1 1000 yr−1 at t ≥ 6 months 79

t = 0: first survey; t ≥ 6 months: second survey.a Female rate of seroconversion significantly higher than male (�2 = 4.4

26.0 0.67 NS32.7 0.89 NS

ere also significantly older than those of intensified onesP ≤ 0.05).

The overall mean age of persons seroconverting in bothlantation types during the study was almost exactly theame as that recorded for the prevalence survey. Althoughhe overall difference was not significant, residents of tradi-ional plantations who seroconverted during the study wereignificantly younger (P ≤ 0.05) than those of intensifiednes in both Salamina and Gramalote. Individuals who sero-onverted between the two visits were aged 3—82 years inraditional plantations and 2—85 years in intensified ones.

There were no marked differences in the prevalence

f Leishmania transmission between males and females inraditional plantations, except for the residents of Toledo.lthough not significant at P ≤ 0.05, the overall preva-

ence of infection for males was slightly higher than that foremales. With regard to intensified plantations, a significant

ed on leishmanin skin test (LST)] by sex among residents ofunicipalities

antation type

aditional Intensified

F M F

.3 24.0 14.6 12.1

.4 81.0 5.0 12.0a

4; P = 0.035).

lomb

oottoasalcdrondfubdLpwpeas

pttrsaltwepwao

4bcfinmd

mLcottc

Coffee-growing practices and Leishmania transmission in Co

difference was found in a single municipality (Pensilvania)and overall prevalence was again higher in males than infemales, although this was not significant.

Differences between the sexes with regard to the inci-dence of Leishmania infection did not show the same patternas prevalence values. No significant difference was seen forresidents of traditional plantations and the overall incidenceof infection was roughly the same for males and females.In intensified plantations significant differences (P ≤ 0.05)were detected for Salamina as well as for the overallincidence. Thus, in intensified plantations a greater serocon-version rate was seen for females during the study period.

4. Discussion

Although the sample size was limited by the number of plan-tations that could be visited on foot during the two visits toeach municipality, the results of this study provide evidencethat Leishmania transmission is indeed lower among resi-dents of intensified plantations than in those of traditionalones. The true difference in prevalence values betweentraditional and intensified plantations is likely to be evengreater, since some of the residents of the latter who gavepositive responses will have become infected with Leishma-nia over 20 years ago, before the Colombian coffee industryadopted a policy of intensification.

The number of plantations under the intensified system ofcultivation has increased gradually since 1984, with growersin some areas more prepared to make the conversion than inothers. Intensified plantations currently account for about70% of the total in Colombia. Most of the residents of theplantations were tenants rather than owners and possessedno record of the dates that conversion between the systemshad occurred.

The principal reason for the lack of a demonstrable dif-ference in seroconversion values between residents of thetwo types of plantation is probably the unexpectedly highprevalence rate, which removed large numbers of individu-als who tested positive in the first survey from the secondround of testing. The number of individuals available for skintesting during the second survey was thus too small to allowa significant difference to be shown at P ≤ 0.05.

The municipalities included in the study belong to twodepartments of Colombia that differ with respect to thesand fly species that transmit Leishmania,14 the parasitesresponsible for ACL and cultural practices that might affectthe degree of human—vector contact. In the Norte deSantander plantations the dominant sand fly species is Lut-zomyia spinicrassa (Morales et al.), an avid human-biter thatenters houses and is a suspected vector of Le. brazilien-sis. This sand fly species is apparently absent from Caldas,where potential vectors of Leishmania in coffee plantationsinclude Lu. hartmanni (Fairchild & Hertig) and Lu. gomezi(Nitzulescu) (both suspected vectors of Le. panamensis) andLu. ovallesi (Ortiz), a suspected vector of Le. braziliensis.15

ACL due to Le. mexicana has also been recorded from the

Eje Cafetero.16 No attempt was made during the presentstudy to identify the parasites responsible for ACL in eitherof the two departments studied.

The differences in transmission rates between intensifiedand traditional plantations could be attributed to a number

rfdpi

ia 1267

f factors, including differences in the sand fly fauna, usef agrochemicals and degree of human—vector contact. Inraditional plantations sand flies could be collected duringhe day on the trunks of shade trees within a few metresf houses as well as among leaf litter and they may havelso rested in animal burrows. In intensified plantationshade trees were absent, leaf litter much less abundantnd the food sources available to small mammals muchess plentiful.14 However large numbers of sand flies wereollected on intensified plantations (both at night and iniurnal resting sites) and the relatively high seroconversionates among residents indicated that a significant amountf Leishmania transmission continues to occur under theew system. Distance between the coffee crop and humanwellings in the plantations visited was never more than aew metres, so that the limited flight range of sand flies17 isnlikely to have been an important factor in the differenceetween human infection rates. Exposure to Leishmaniaoes not necessarily result in clinical disease and theST+ve individuals who tested in the present study includedeople who would eventually develop ACL and others whoould remain healthy. Estimates of the ratio between therevalences of infection and disease vary although Weiglet al.18 found that on the Colombian Pacific coast 9.25 timess many people were infected by Leishmania as showedymptoms of ACL.

The fact that there was no significant difference inrevalence values between males and females in eitherype of plantation supports our view that most Leishmaniaransmission in these situations occurs in and around housesather than in the coffee crop itself. Large numbers ofeveral vector species can be collected inside houses14

nd although sand flies will bite during the day when lightevels are low or if disturbed from diurnal resting sites,his is a relatively rare occurrence. However risk to malesould be expected to be somewhat greater because theyxpose more skin to sand fly bites and are more likely toass through forested areas after dark when returning fromork or hunting. The significantly higher incidence seenmong females in intensified plantations is further evidencef intra- or peridomiciliary transmission.

Since 1970 intensification has transformed about00 000 ha of traditional plantations in which 1000 arabicaushes produced 600 kg of coffee/ha into colombia mono-ultures producing seven times as much coffee at densitiesve times as high.19 Reduced Leishmania transmission canow be added to the benefits of intensification, although thisust be weighed against detrimental environmental effectsue to increased use of agrochemicals and soil depletion.

In Colombia ACL is a notifiable disease and futureonitoring of the effect of coffee-growing practices on

eishmania transmission could be significantly improved byhanges in the data collection forms used by the Ministryf Health. Patients reporting to local health services forreatment are asked a number of questions but at presenthere is no way of determining how many infections wereontracted in coffee plantations, since the occupations of

ural inhabitants are categorised as ‘agricultural worker’or men, ‘housewife’ for women and ‘student’ for chil-ren below working age. As shown by the results of theresent study, both sexes and all age groups may thus benfected and ACL is not solely an occupational disease.

1

QtwosCtcgsntfida

Acpim

AfCaNbdAtasAv

F9

C

EamEbtattsa

Rvas. Centro Nacional de Investigaciones ‘‘Pedro Uribe Mejia’’

268

uestions on whether the patient works or resides on a plan-ation and if so, type of plantation (shaded or unshaded)ould provide valuable information on the long-term effectsf intensification on ACL prevalence. King et al.20 con-tructed a predictive logistic regression model of ACL inolombia, based on elevation and land use data (the lat-er obtained from satellite images). They found that landover had greater predictive power than elevation and sug-ested that their approach could be improved by usingatellite imagery with greater spatial resolution. An alter-ative would be to improve information exchange betweenhe health authorities and FEDERACAFE, the Colombian cof-ee growers’ association. This could include improvementsn data gathering from ACL cases, as well as incorporatingetails of leishmaniasis and awareness of its transmissionmong respondents in every future coffee-growers’ census.

uthors’ contributions: All authors participated in studyonception and design, data collection and analysis, and thereparation of the paper. BA revised the paper critically forntellectual content. All authors read and approved the finalanuscript. BA is guarantor of the paper.

cknowledgements: The logistic support supplied by theollowing people is gratefully acknowledged: Idelfonsoepeda of the Ministério de Salud, Bogotá; Fiorentino Célisnd Hector Anaya of the Servício Seccional de Salud deorte de Santander; Carlos Arturo Feho Moncada, Hum-erto Bedoya and Julian Montoya of the Servício Seccionale Salud de Caldas; Ivan Dario Velez of the University ofntioquia, Medellín and Felipe Guhl of the University ofhe Andes, Bogotá. We are particularly grateful for thessistance in the field provided by the promotoras dealud and other health workers of the municipalities ofguadas (Arma), Bochalema, Gramalote, Labateca, Pensil-ania (Pueblo Nuevo and San Daniel), Salamina and Toledo.

unding: This study was funded by WHO/TDR grant no.403334.

onflicts of interest: None declared.

thical approval: The study was carried out with the fullpproval of the Colombian Ministry of Health and the depart-ental health authorities in Caldas and Norte de Santander.

thical clearance to carry out skin testing was providedy the local hospital directors in each of the municipali-ies visited and local health workers (promotoras de salud)ccompanied the authors throughout, supervising applica-ion of the antigen and often applying it to participantshemselves. All study participants or their legal guardiansigned a consent form before the leishmanin skin test waspplied.

eferences

1. Anonymous. La rebelión cafetera. Cambio 1999;333:18—28.2. Young DG, Morales A, Kreutzer RD, Alexander JB, Corredor A,

Tesh RB. Isolations of Leishmania braziliensis (Kinetoplastida:

B. Alexander et al.

Trypanosomatidae) from cryopreserved Colombian sand flies(Diptera: Psychodidae). J Med Entomol 1987;24:587—9.

3. Alexander B, Ferro C, Young DG, Morales A, Tesh RB. Ecol-ogy of phlebotomine sand flies in a focus of Leishmaniabraziliensis in Northeastern Colombia. Mem Inst Oswaldo Cruz1992;87:387—95.

4. Alexander B, Lozano C, Barker DC, McCann SHE, Adler GH.Detection of Leishmania (Viannia) braziliensis complex in wildmammals from Colombian coffee plantations by PCR and DNAhybridization. Acta Trop 1998;69:41—50.

5. Reithinger R, Cadena H, Davies CL, Alexander B. The fungusBeauveria bassiana as a potential control agent of phle-botomine sand flies (Diptera: Psychodidae) in Colombian coffeeplantations. J Invertebr Pathol 1997;70:131—5.

6. Scorza JV, Rojas E. Caficultura y leishmaniasis tegumen-taria en Venezuela. Bol Dir Malariol San Amb 1988;28:114—27.

7. Scorza JV, Castillo L, Rezzano S, Márquez M, Márquez JC.El papel del cafeto en la endemicidad de la leishmaniasiscutánea en Venezuela. Bol Dir Malariol San Amb 1985;25:82—8.

8. Warburg A, Montoya-Lerma J, Jaramillo C, Cruz-Ruiz AL, Ostro-vska K. Leishmaniasis vector potential of Lutzomyia spp.in Colombian coffee plantations. Med Vet Entomol 1991;5:9—16.

9. Montoya-Lerma J. Transmisión de leishmaniasis tegumentariaen zonas cafeteras de Colombia. Entomólogo 1990;66:1—2.

10. Montoya-Lerma J, Jaramillo C, Palma G, Gómez T, Segura I,Travi B. Report of an epidemic outbreak of cutaneous leishma-niasis in a coffee-growing area of Colombia. Mem Inst OswaldoCruz 1990;85:119—21.

11. Alexander B, Barbosa de Oliveira E, Haigh E, Leal de AlmeidaL. Transmission of Leishmania in coffee plantations of MinasGerais, Brazil. Mem Inst Oswaldo Cruz 2002;97:627—30.

12. WHO. The leishmaniases. Geneva: World Health Organization;1990. Technical Report Series No. 701.

13. Palma G, Gutierrez Y. Laboratory diagnosis of Leishmania. ClinLab Med 1991;11:909—22.

14. Alexander B, Agudelo LA, Navarro F, Ruiz F, Molina J, Aguilera G,Quinones ML. Phlebotomine sandflies and leishmaniasis risks inColombian coffee plantations under two systems of cultivation.Med Vet Entomol 2001;15:364—73.

15. Young DG, Duncan MA. Guide to the identification and geo-graphic distribution of Lutzomyia sand flies in Mexico, the WestIndies, Central and South America (Diptera: Psychodidae). MemAm Entomol Inst 1994;54:1—881.

16. Corredor A, Kreutzer RD, Tesh RB, Boshell J, Palau MT, CaceresE, et al. Distribution and etiology of leishmaniasis in Colombia.Am J Trop Med Hyg 1990;42:206—14.

17. Alexander JB. Dispersal of phlebotomine sand flies (Diptera:Psychodidae) in Colombian coffee plantations. J Med Entomol1987;24:552—8.

18. Weigle KA, Santrich C, Martinez F, Valderrama L, SaraviaNG. Epidemiology of cutaneous leishmaniasis in Colombia: alongitudinal study of the natural history, prevalence, and inci-dence of infection and clinical manifestations. J Infect Dis1993;168:699—714.

19. Valenzuela Samper G. Investigación y desarrollo cafetero. In:50 Anos de Cenicafé 1938—988: Conferencias Conmemorati-

Chinchina, Colombia; 1990, pp. 3—8.20. King RJ, Campbell-Lendrum DH, Davies CR. Predicting geo-

graphic variation in cutaneous leishmaniasis, Colombia. EmergInfect Dis 2004;10:598—607.