epidemiology of trypanosoma cruzi in the

Am. J. Trop. Med. Hyg., 33(6), 1984, pp. 1084-1095 Copyright © 1984 by The American Society of Tropical Medicine and Hygiene

EPIDEMIOLOGY OF TRYPANOSOMA CRUZI IN THE ORIENTAL PLAINS OF COLOMBIA * A. D'ALESSANDRO, PABLO BARRETO, NANCY SARAVIA, AND MAURICIO BARRETO The International Collaboration in Injectious Diseases Research Program, Centro Internacional de Investigaciones Médicas, Tulane University-COLCIENCIAS, Apartado Aéreo 5390, Cali, Colombia, and Facultad de Salud, Universidad del Valle, Cali, Colombia

Abstract. Epidemiologic studies to define the domiciliary and extradomiciliary transmission cycles of Trypanosoma cruzi and Trypanosoma rangeli in the Oriental Plains of Colombia were conducted in the gallery forests near Carimagua and El Porvenir. Onehundred and seven palm trees belonging to nine genera were examined; triatomines were found in only three palm species, the lea ves of which are locally used for roof thatching: 2/29 Maximiliana elegans, 1/7 Afauritia flexuosa and 7/7 Scheelea sp. Bugs were also found in 5/14 hollow Afauritia inhabited by bats, 4/21 bird nests and 1/4 armadillo burrows. Five species oftriatomines were collected: Rhodnius prolixus was the most abundant, 192 of the total 207 (92%) collected; the bugs were found in iJaximiliana and Mauritia but especially in Scheelea, and 8% were infected with T. cruzi and T. rangeli; Cavenicola pilosa and Triatoma maculata were found associated with bats; Psammolestes arthuri and Panstrongylus lignarius with bird nests and Panstrongylus geniculatus with armadillos. Although triatomine colonies were not found in human dwellings, fiying adults of R. prolixus occasionally reached houses by their own locomotion and fed on man, but did not become established. Only 12 of 199 persons (6%) tested serologically were reactors to T. cruzi antigens and all12 had lived in areas ofdomiciliary transmission elsewhere in the country, indicating that domiciliary transmission is not occurring in this region. Whether the presence of domiciliary R. prolixus in houses loeated in the ecologically altered piedmont of the oriental plains, a known area of domiciliary transmission of T. cruzi, is due to importation of domiciliary bugs from endemic areas or to the domiciliarization ofwi1d R. prolixus remains to be determined. The two American trypanosomes ofman, Trypanosoma cruzi and Trypanosoma rangeli, are known to be present in Colombia and their distribution in the country has recently been summarized and updated by D'Alessandro et al.l-2 The country is divided into natural regions f01lowing the water distribution system (Fig. 1). Endemic regions for man are: the Catatumbo River basin, in the northwest, the Magdalena River valley in the central region, and the eastern slopes Accepted 15 May 1984. * Supported in part by the Tulane University-COLCIENCIAS, Centro Internacional de Investigaciones Médicas, CIDEIM; by grant AI-10050 from the National Institute of A11ergy and InfectiousDiseases,NIH, USPHS, and by grant 97335-3-01-82 from the Fondo Colombiano de Investigaciones Científicas, COLCIENCIAS. Address reprint requests to: Dr. Antonio D'Alessandro, Department of Tropical Medicine, Tulane University MedicalCenter, 1430Tulane Avenue, New Orleans, Louisiana 70112.

and piedmont ofthe Oriental Cordillera. To the east, lie the oriental plains, formed by the Orinoco River basin, and limited to the south by the Guaviare River and the beginning ofthe Amazon River basin. Whether or not domiciliary transmission occurs in the Orinoco and Amazon River basins of Colombia, has not been established. The Atlantic coast region is potentially endemic due to the presence of small apparently imported domiciliary foei of Rhodnius prolixus. In the remaining regions-Cauca River Valley, Andean Central Massiff and Pacific Coast-as far as is known, transmission to man does not occur due to the absence of domiciliary triatomines. However, in every region below 2,700 m, animals and extradomiciliary vectors have been found infected with T. cruzi or T. rangeli, or both. In the human endemic regions, where most known T. cruzi infections and Chagas' disease cases are found, the main domiciliary vector of both trypanosomes is R. prolixus. The objective ofthis study was to characterize the zoonotic cyde of T. cruzi and to determine

1084

EPIDEMIOLOGY

OF TRYPANOSOMA

~I

I ~)' ¡VI ¡U ¡

'~

!

\

¡

I

¡

(

~-----l

I

CRUZ/IN

COLOMBIA

1085

(range 14-35°C). The elevation is 150-157 m. The area is characterized by interfiuvial savannahs, and tropical dry gallery forests of deciduous hardwood and palms. The field work was carried out from July 1978 to December 1979, but wild animals were collected since 1972.

¡

Wild animal studies P.,CfF/C OCU.H

"-~"-'"

FIGURE

l.

Map showing natural regions of Co-

lombia.

its importance and relationship to human infection and human disease. We investigated the epidemiology of the sylvatic, peridomiciliary and domiciliary cycles of transmission in different areas. In this report we present results of studies carried out in the area of Carimagua and in El Porvenir, in the oriental plains ofColombia. The trypanosomes iso1ated from triatomine vectors and wild animals were characterized morphologically and biologically. Their intrinsic biochemistry, determined using isoenzymes, will be published separately. MATERIALS AND METHODS

Studyarea The study area included the Carimagua Experimental Station of the Instituto Colombiano Agropecuario (lCA) and neighboring ranches, including areas in and around El Porvenir, a small town on the Meta River.3 The annua1 rainfall is 1,800 to 2,000 mm, distributed from April to November. The mean annual temperature is 27°C

Anima1s were trapped or shot during field activities. Their study included the following procedures: 1) hemoculture in Tobie et al.'s medium;4 2) thick b100d smears stained with Wright' s and then with Giemsa stains (pH 7); 3) xenodiagnosis using 10-40 fourth instar nymphs of Rhodnius prolixus. Insects were subsequently fed on mice, examined after each molt and dissected 40-100 days 1ater using procedures previously described.5 Feces and gut contents showing fiagellates were inoculated intraperitoneally into mice. These mice, as well as those used to feed infected bugs, were bled from the heart at day 8, and the bood was cultured in Tobie's medium. Triatomine studies Triatomines were collected in the following sites: 1) palm trees. The leaves of felled palm trees were cut off onto a piece of white canvas and examined for triatomines; also hollow palm trees, with or without bat inhabitants, were felled, split open and triatomines searched for; 2) bird nests; 3) terrestrial mammal burrows; 4) thatchedroofed human dwellings and chicken coops. Preserved nymph and adult triatomine bugs were shown to local people and records were kept of (a) whether such insects were recognized and called by the local name (pito), and (b) whether or not they were permanent residents or visitors of the present or previous dwellings. After the interviews, search for triatomines was carried out in El Porvenir as well as in several houses on both sides of the Meta River. At least two experienced assistants performed the search for the insects or their remains, without keeping a record of the time involved. The thatched roof houses are usually constructed ofwood trunks, the walls and partitions being of split bamboo. The village and other houses visited were surrounded by gallery forests, with cleared areas ranging from a few to 100-200 m in length. DDT spraying had been carried out about every 6 months for at least 2-3 years before the study began.

1086

D'ALESSANDRO

In addition, attempts were made to collect triatomines using light traps, both in the gallery forest and outside our small station in the town of El Porvenir. The 15 watt ultraviolet light trap employed was similar to the model used by Whitlaw and Chaniotis in panama.6 The source of the white light was a 500 cd Coleman lamp placed on a white canvas, either on the ground or on the hood of a jeep. The triatomines collected were studied repeatedly using the same protocol mentioned under wild animal studies. Attempts were made to adapt the different triatomine species collected to laboratory conditions, as well as to establish colonies of them. ldentification of blood meals Gut contents offully or partially engorged bugs were expressed onto filter paper, dried, and stored at 4°C in screw cap vials with dessicant until tested. Blood spots were cut out and eluted in physiological saline ovemight at 4°C. The volume of saline used varied according to the size and color intensity of the blood spot. Antisera for precipitin testing were produced in rabbits. Broad spectmm screening antisera of birds, mammals, and reptiles-amphibians were obtained after multiple inoculations with a pool of sera from a locally obtainable variety of species, each representing a distinct phylogenetic order. Antisera that were more restricted in specificity were generated using a single species or pool of closely related species sera as immunogen. Absorption with lyophilized sera from cross-reacting species was necessary in some instances, to attain specificity to the leve! of order. Immunodiffusion (ID) was performed in agar by double diffusion (Ouchterlony). Following an initial screen test using anti-mammal, anti-bird, and anti-reptile amphibian, eluted blood samples were tested with the following antisera: primate including human, camivore, rodent, edentate, bat, horse, cow, opossum, rabbit, Cuniculus paca, and chicken. Maintenance and characterization of culture isolations Isolates obtained from wild animals and from inoculated mice were kept at 27°C in Tobie's medium and transferred monthly. Each isolate was studied by the following procedures: 1) morphological characteristics of culture forms in

ET AL.

Giemsa-stained preparations; 2) hemocele inoculation of flagellates from cultures using procedures described previously.5 Hemolymph samples were examined on day 5 after inoculation, and if flagellates were not observed the insect was dissected at day 10 and its organs searched for flagellates. However, if parasites were present the insects were examined periodically to assess invasion ofthe salivary glands; 3) Vero (green monkey kidney) cell cultures were inoculated with washed culture forms ofthe isolates (106 flagellates per culture tube). The cell monolayers grown on coverslips were fixed and Giemsa stained on day 6 after inoculation and examined for amastigotes; 4) infectivity for mice was evaluated by intraperitoneal inoculation of four to six 10 g male outbred albino mice with 1-6 x 105 flagellates from 15 day cultures. Infection was verified by hemoculture on day 8. Flagellate identification The presence of amastigotes in tissue culture biologically defined T. cruzi (Tcb) and the characteristic morphology in blood, in culture and/ or in triatomines morphologically identified T. cruzi-like (Tcm) isolates. The presence of metatrypomastigotes in the bugs' salivary glands and/ or bite transmission to mice biologically identified T. rangeli (Trb), while the morphology in triatomines identified T. rangeli-like flagellates (Trm). From bats, T. cruzi-like isolates infecting mice were called T. cruzi, b (= T. cruzi cruzi Baker et al., 1978);7 and on the basis ofthis biologic behavior isolates not infecting mice were denominated T. vespertilionis, b (=T. cruzi marinkellei Baker et al. 7). Trypanosoma (Schizotrypanum) biological and morphological (TSb, TSm) identify, respectively, flagellates assessed only by tissue culture for the production of amastigotes or by blood smears, showing a T. cruzi-like morphology. Serology Human semm aliquots were derived from blood samples obtained on a voluntary basis during a morbidity survey carried out in November 1979 in El Porvenir and among workers of the ICA Station, most ofwhom lived in or had lived in the surrounding areas before they started working in Carimagua. All 36 inhabited houses in El Porvenir were visited.

EPIDEMIOLOGY

OF TRYPANOSOMA

TABLE

CRUZI IN COLOMBIA

1087

1

Characteristicsof the human population surveyed in El Porvenir-Carimaguaarea: age and sex distribution, recognitionof triatomine bugs, averagetime of residencein the area, and results of serologicalstudiesfor T. cruzi

---

1/16 Females No Males Yes Males T Total otal x time in area (yrs) 1/12 15 4 4 11 Recognition of 22 18 4.3 1 17 3Females 2 13 29 6.1 43 15.11/10 15.7 6 O 39 3 514 128 21 6.7 12/199 76 83 .4 6 5.3 O 7 18 153 0/50/3 3/65 3/41 67 6 1 0/4 0/4 0/17 0/16 0/2 3.0 0/1 1/1310 1/1031 2/31 7.6 1/6 10/132 3 0 /48 /3 6 29 19 53 3 64 .3 83 35 0/32 0 3/13 /16 30 55 229 33 13 10 4/26 2/67 100 6.0 17triatominesResults of serological studies·

* No. positive/no. examined.

t Case histories

were 10S1.

Blood samples were obtained from 68 and 119 persons (10 years ofage or older) ofthe censused population of 188 and 280 in El Porvenir and Carimagua, respectively. The sex and age distributions, together with other pertinent information are presented in Table l. Indirect hemagg1utination (IHA), complement fixation (CF), and immunofluorescence (lF) were carried out at the Instituto Nacional de Diagnóstico e Investigación de la Enfermedad de Chagas in Buenos Aires, Argentina, and IHA, as part of another study, at CDC, Atlanta. Persons with one or more positive tests for T. cruzi antibodies were considered reactors. RESULTS

Sylvatic ecology of potential vectors One hundred and seven palm trees, belonging to nine genera and 10 species were examined: 1 Acrocomia sp. (corozo montañero); 2 Acrocomia sp. (corozo sabanero); 3 Euterpe precatoria (manaca); 28 Mauritia flexuosa (moriche)-2l dead and hollow; 1 Mauritiella sp. (moriche macho); 29 Maximiliana elegans (cucurita); 1 Oenocarpus minor (pusuy); 7 Scheelea sp. (palma real); 26 Socrate exorixa (chuapo); 9 Syagrus inajai (churrubay). Triatomines were located at the base of the leaves of only three species of palm trees: 2/29 Maximiliana, 1/7 live Mauritia and 7/7 Scheelea. These species, the most frondose, are used in this locality for roof thatching. Triatomines

were also found in 5/14 dead and hollow Mauritia inhabited by bats, 4/21 bird nests (3/3 nests of the Plainfronted Thombird, Phacellodomus rufifrons and 1/1 Crested Cara-cara, Polyborus plancus), and 1/9 burrows (1/4 burrows of armadillo, Dasypus sp., 0/2 pacas, Cuniculus paca, and 0/3 undetermined rodents). The species and number oftriatomine vectors collected, distribution by collection sites, source of blood meal identified, and trypanosome infection encountered are shown in Table 2. In sylvatic sites, six species oftriatomine bugs were collected: Rhodnius prolixus, Cavernicola pilosa, Psammolestes arthuri, Triatoma maculata,' Panstrongylus geniculatus and Panstrongylus lignarius. R. prolixus, the main domiciliary vector ofboth T. cruzi and T. rangeli in Colombia and Venezuela, was the most abundant. We collected 207 R. prolixus in three species of palm trees: Mauritia, Maximiliana and Scheelea, the latter being the most frequently infested (7/7) and where most of the insects were found (192 of 207, 92%). The average number of R. prolixus per tree was 27. In contrast, only two to six bugs per tree were collected in Mauritia and Maximiliana. Forty-six C. pilosa were collected in 4/14 hollow Mauritia inhabited by bats, however, many other bugs escaped. P. arthuri was found in all three thombird nests examined, an average of 30 bugs per nest; in addition, 600 eggs were also found, some parasitized by the wasp Telenomus capito. Fourteen T. maculata were found in two bat-inhabited hollow Mauritia and in a Scheelea.

o00 ....-

00

-----

-

-

291 426946 192 tested* 5No. 25 Id+/no. Instars no. posi- r' Total Habitat tals M43/2 infect.lno. examined Eggs m RlA 24 43Mauritia, Trypanosome Blood meal speciest type 112 t) No.bugs 600 0/78 bat 9114 46/4 2/28 3/3 12/1 11 262 1192/7 3F2~ 1/2 836473 2sp. 13 10/155 bird 19/20 10 bird 24 5/6 bird-bat Mauritia 2117/21 2/13 4251 bird-RlA 12bird 10 8Trb Tcb Thombird Maximiliana Tcm; 191/3 Trcm; sp sp. bird-RlM Scheelea sp. 1 Tcb; Trm No.bugs/ 1 undetermined nir-Carimagua by instar, habitat, blood meal type, and trypanosome infection C/J O r< >

m

No. collected

..., C/J

TABLE2

;;t)

~

----------

231514 852311/1 Mauritia Scheelea Mauritia,

-- - ~n~~ 0/37 212 5/8 .....•

5

O te S

93Porvenir 211 tested* 22 330 No. 5 Tcb M Id+/no. Instars examined infect/no. Habitat tats 25 tI1 F:>;: No. s:: 712Total Eggs Trypanosome bugs Bloodmea!type 2516 speciest 5/1 10/14 234El El Armadillo 24 2/1 undetermined no. posi- O 1Obat 31Camatarigua Provenir 12/3 human Tcb 1412 29/47 2/7 :;J R/A 34 3/1 ." bugs/ >r1 t""' *tI1 Tr, RI A, Trypanosoma reptile/amphibian. Te, Trypanosoma cruzi; b,forms biologically; m, morphologically. §s::Except Reported farpreviously.2 two liverangeli; 4th instars, the wereNo. exuviae all entangled Q other immature

M aximiliana, tZ ;; ~1 Tcm; Trb Continued 11

TABLE 2 in a spider web.

O

~

¡g

•...•

O 00 '-O

1090

D'ALESSANDRO

Finally, a few P. geniculatus and P. lignarius were collected, the first in an armadillo burrow and the second species in a M auritia and a toucan nest. Vectors in domiciliary habitats Although domiciliary colonies were not found in El Porvenir, three species of flying adult triatomines were collected over several months in human dwellings: 58 R. prolixus and one each of T. maculata and E. mucronatus. During March 1978 (end of the dry season) the zinc roof of our station in El Porvenir was changed for a thatched one, using 4,600 locally obtained Mauritia leaves. While the roof was being installed and thereafter, the number of R. prolixus seen in the house increased. The number of adults, and the only instars seen, were distributed as follows: July 1977 through February 1978, none; March, 19; April, 9; May, 18; June, July, August, 1-2/ month; and September, October, November, none. To determine whether the bugs could have been introduced with the leaves, 842leaves from Mauritia and 600 from Maximiliana palm trees (cut near El Porvenir for use in roof thatching) were examined carefully at the time of roofing, about 15 days after being cut, but neither triatomines nor their eggs were seen. In one house examined in Camatarigua, downriver from El Porvenir, R. prolixus had apparently become or attempted to become established: two fourth-instar nymphs and 33 exuviae of2nd, 3rd and 4th instar nymphs were collected, all entangled in a spider web. In addition, 55 adults and two 4th instar nymphs were obtained by the owner ofthe house over a period of several weeks. Light trapping The attempts to collect triatomines using light traps (white and ultraviolet) both in the gallery forest and outside the El Porvenir Station totaled 48 hours. Although a variety ofinsects were collected, triatomines were not among them. Serology of the human population The results of the serological tests carried out on the samples obtained from the human population under study indicated only four reactions out of 68 persons in El Porvenir and eight of 119

ET AL.

in Carimagua. IF showed the highest rate ofpositivity: 11/11 vs. 7/11 in CF and 7/12 in HA. Eight ofthe asymptomatic reactors had lived in known endemic areas of domiciliary T. cruzi transmission and stated that their houses (in those places) had been infested with triatomines. However, one person may have acquired the infection in El Pauto, Casanare (not far from El Porvenir), but no records were available as to other places where he had lived (Table 1). Reservoir studies Ofthe 737 wild animals examined (Table 3), trypanosome infections were observed in edentates, 33%; marsupial s, 20%; camivores, 19%; rodents, 9%; and bats, 6%. The overall rate of infection was 12%. The most frequently infected species were the opossum, Didelphis marsupialis, 43% (27/62; 18 T. cruzi, 2 T. rangeli, 3 both); the paca, Cuniculus paca, 13% (10/74, all T. cruzz); and the spiny rat, Proechimys spp., 11% (24/224; 15 T. cruzi, 1 T. rangeli, 1 both). DISCUSSION

Triatomine bugs are of sylvatic origin, and their degree of adaptation to human dwellings varies. Some species live in close contact with wild animals and never or seldom have contact with mano Others, common visitors to human dwellings, are sometimes able to breed in the new environment. Finally, a few have become so well adapted to the domestic environment that it is difficult to find them in their original habitats.8-9 Among the latter are T. infestans and R. prolixus, the two most important vectors of human trypanosomiasis in large areas ofCentral apd South America. These two species were known to live in human dwellings and peridomestic sites such as chicken and pigeon coops, but the occurrence outside these sites was considered exceptional. Gamboa reported finding all stages of R. prolixus in 7 of 30 palm species of about 500 palm trees, representing practically every Venezuelan state, often far from human dwellings. 10-12 Almost every Scheelea humboldtiana examined was found to be inhabited by up to 19 R. prolixus, some of which were infected with T. cruzi, or T. rangeli, or both species. R. prolixus eggs were also found attached to the leaves. It has been generally accepted that beca use leaves of this palm are frequently used in roofthatching, R. prolixus could be transported passively to the newly constructed

----- - --

Trm t21 bTSb Both Trb Tcb spp 7TSm 1 13 17 4225321124Tcm 10

t Theileria-like.

examined

";.,........,..

Trypanosome

0/19 0/7 2/7 1120 0/2 92/757 24/224 27/62 5/5 0/6 0/2 3/11 2/6 0/32 3/7 1115 114 1110 112 0/1 10/74 0/1 0/43 0/29 0/37 5/57 0/33 2/2 118 113 1111

(12.0%) ons (Schizotrypanum); in mammalsfrom b, biologieal; Carimagua-EI ID, morphologieal. Porvenir Didelphis Dasypus Cuniculus Odocaileus Cerdocyon kappleri marsupialis paca thous virginianus Artibeus jamaicensis No. pos.lno.

infeetions*

~f

1091

1092

D' ALESSANDRO ET AL.

houses. In addition, the presence oflarge numbers of R. prolixus in nests of four species of stork-like birds and less frequently in those of other birds, as well as the observation of eggs and young nymphs of this insect on the feathers of the large Jabiru mycteria, further suggested that birds can passively transport R. prolixus during migration. More recently, other reports fromVenezuela confirm those observations and extend the number of species of palm trees infested.13-15 In the Canal Zone of Panama, the infestation rate of Scheelea zonensis with Rhodnius pallescens was about 100%. A direct relationship between palm infestation and domiciliary colonization of this insect, the main human vector in the country, was suggested.6 Until recently the question ofwhether or not extradomiciliary R. prolixus were present in Colombia was unsettled. 16-17However, its presence in palm trees and nests is now well established by the observations of D' Alessandro et al.2 and those presented here. It is evident that R. prolixus occurs frequently in at least three species of palm trees in the Oriental Plains, not only in the area of Carimagua and El Porvenir but also in the piedmont around the city ofVillavicencio, Meta, and probably elsewhere (unpublished observations). The infested palm trees are those with most abundant foliage, which permits the sheltering of mammals, birds, reptiles and amphibians that pro vide a source of blood to the wild triatomine population. The "palma real," Scheelea sp., has the thickest foliage and its leaves are the best and most commonly used for roof thatching. Because of this, they are the most valuable commercially and it is harder to obtain permission for cutting. All seven in our series were infested with R. prolixus, which were present in higher numbers in these than in the other two palm species, Mauritia and M aximiliana. The immunodiffusion tests showed that the bugs had fed mostly on birds and on reptiles and amphibians, but also on rodents and one on a primate. The T. cruzi and T. rangeli infection rate was 8%, indicating that they had access to and at some time had fed on infected mammals. Of the other triatomine species found in our study, C. pilosa had been previously reported in the Oriental, Magdalena and Cauca River Valley region of Colombia. Interestingly some of the bugs we captured had fed on birds in addition

to bats. On the other hand, P. arthuri had only been known to inhabit thornbird nests in Venezuela. A large proportion of the eggs collected were parasitized by the wasp Telenomus capito, which also appears to have been previously unreported in Colombia.18 T. maculata has been reported in the Atlantic coast region as a domiciliary and peridomiciliary species, but its association with bats had not been previously observed. Of the two species of Panstrongylus collected, P. geniculatus has been recorded in every Colombian region and its association with armadillo s is well established. On the other hand, P. lignarius had been collected in palm trees in this area and only recently been reported in Co10mbia.2 These five triatomine species are known to be naturally infected with T. cruzi in sylvatic environments and, therefore, playa role in the zoonotic transmission of this flagellate. Only T. maculata has been implicated in human transmission. Our limited experience (48 hours) with light traps was unproductive. In Venezuela, Tonn et al. 19 using the Sjogren and Ryckman type trap, captured low numbers of four species of triatomines including R. prolixus, which was the most frequently attracted (ca 0.024/hour). In Panama,6 four of the five triatomine species present in the country were also collected in appreciable numbers using light traps. Whether our failure was due to the low light intensity or to the type oftrap used, short duration oftrapping, low density of the sylvatic triatomines present in the area, or a combination of these is not known. Gómez-Núñez stated that in the Venezuelan Orinoco basin, huts long inhabited by man were triatomine- free although they were located among R. prolixus-infested palm trees.20 To explain this observation he suggested that sylvatic and domiciliary R. prolixus were separate species, but this hypothesis was not substantiated in crossbreeding experiments; nonetheless, it was concluded that the morphological differences observed in the two populations probably were an expression of initial genetic differentiation. Similar views have been expressed with regard to P. megistus in Brazil.21 In addition to the passive transfer of sylvatic R. prolixus to human dwellings (on palm leaves or household goods) their arrival by active 10comotion has also been considered.22 GómezNúñez,23 using C060-tagged R. prolixus, observed that their movement was induced mainly

EPIDEMIOLOGY

OF TRYPANOSOMA

by starvation. After releasing R. prolixus in palms and huts, emigration from palms to houses was greater than between palms or between houses, and house to palm movement was not observed. No significant differences in behavior were noted between insects originally captured on palms and those captured indoors. In Gómez-Núñez experiments,23 dispersal did not exceed 15 m outdoors and 4 m indoors; on the other hand, the observations ofRabinovich et al. 24 based on capture-recapture of domiciliary R. prolixus indicated that both nymphs and adults were recaptured at distances of 100 to 500 m. We have seen fiying sylvatic R. prolixus arrive at our Station in much the same fashion as other sylvatic species oftriatomines were observed to do in the Amazon region of Brazil.25 In both instances, because of the absence of colonies in human dwellings domiciliary transmission does not occur, as proven by the almost complete lack of T. cruzi serological reactors among the human inhabitants. The few human infections seen by us very probably took place while these persons lived in endemic areas of domiciliary transmission (in dwellings with domiciliary triatomines, probably R. prolixus), or in a few instances accidentally by visiting triatomines. The actual mode of entry to the house may be both aerial and terrestrial. Once inside, the insects may be eaten by predators (fowl, rats, spiders), they may die in the house of old age or be killed by the inhabitants; or they may emigrate back to the sylvatic or some other environment, which is less probable. Apparently only rarely (our observations in Camatarigua, a place not different from and close to El Porvenir) has R. prolixus been found colonizing human dwellings in the gallery forest of the Oriental Plains far from the piedmont, and whether the insects were originally domiciliary or sylvatic in origin has not been established. Unpublished serological surveys carried out by us in the piedmont areas near Villavicencio demonstrated high T. cruzi infection rates (municipio Cumaral, Meta, 11/34, 32%; municipio Medina, Cundinamarca, 8/31, 26%). On the other hand, we have observed low rates of serological reactors in the areas of the Amazon River Basin (municipio Leticia, Amazonas, 0/125; municipio Puerto Inírida, Guainía, 5/47, 11%, and municipio Puerto Leguízamo, Putumayo, 101 77,13%). The people interviewedin the Amazon River Basin, provided the same response to in-

CRUZI IN COLOMBIA

1093

quiry as that obtained in El Porvenir-Carimagua. Some of the people recognized triatomine bugs as visitors to dwellings, and many had been living in known areas of transmission outside the basin. In the Puerto Inírida area, the Indian population reported being bitten by triatomines while collecting fibers from the chique-chique palm tree. The question of what factors are involved in the colonization of a house by R. prolixus is of the foremost importance. Passive transfer of R. prolixus from sylvatic to human dwellings with the palm leaves used for thatching has be en suggested,10-12although we were unable to substantiate this view. At least in our study area this mechanism seems unlikely because R. prolixus is usually located at the base of the palm leaves, which remain attached to the tree; furthermore, before the cut leaves are used for thatching, they are dried in piles for 2-3 weeks, during which time triatomines should leave in search ofblood sources unless vertebrates use the site as a refuge or for nesting. The unavailability ofrefuge and blood sources in the wild environment, as a result of the irrational colonization and exploitation of land and the consequent destruction ofthe ecological balance, together with the poor quality ofthe human dwellings built at the sites, are considered to be the two most important factors infiuencing domiciliary colonization by triatomines.26-27 This has been observed by Forattini et al.28 in regions of Brazil where the sylvatic population of the usually domiciliary Panstrongylus rnegistus is showing signs of adaptation to human dwellings in areas being transformed into agricultura! zones. According to Forattini et alY some triatomine populations living in the small remnants of original forest have a stronger ecological adaptability than others living in large unchanged forests, the latter being less prone to adapt to human dwellings. In the Amazon basin of Brazil, Lainson et al. 25 have reported the presence of a sylvatic cycle of T. cruzi transmission with occasional adult triatomines visiting homes located near the forest. The few human infections observed in the region were very probably due to these visits. It was recognized, however, that because of the rapid development of the area, Chagas' disease may become endemic if sylvatic triatomines become adapted to human dwellings or, more probably, if already domiciliated bugs are imported by the immigrants. Our observations in the gallery for-

1094

D'ALESSANDRO

est of the Colombian Orinoco basin are similar although the most frequent wild triatomine is R. prolixus, one of the most domiciliary species known. It is the main human vector of T. cruzi and T. rangeli in Colombia, inc1uding the piedmont ofthe eastern Andean range and beginning ofthe oriental plains. It is in the piedmont region where the original forest has been c1eared for agricultural use that R. prolixus is responsible for domiciliary transmission of Chagas' disease. Whether the domicilization was due to importation of R. prolixus from endemic areas of human infection or to the adaptation of the wild population cannot yet be established. Attempts are being made to biochemically distinguish wild vs. domiciliary R. prolixus and this information, together with the definition of the zymodeme type ofhuman T. cruzi in this area, may provide a means of achieving a better understanding of the epidemiology of T. éruzi in the region. ACKNOWLEDGMENTS

We wish to express our thanks to Drs. Eric Wells, Pedro Sánchez, James and Joyce Spain, Gustavo Morales, Eduardo Aicardi and Fernando Bernal, Centro Internacional de Agricultura Tropical, CIAT, Palmira and to Dr. Guillermo Matéus, Instituto Colombiano Agropecuario, ICA, for the logistic support provided while working in Carimagua and El Porvenir; to Darío Angel, CIAT, and Drs. Stephen Collett and Jon Yates, Tulane, CIDEIM Research Associates, for providing blood samples from some of the animals investigated during the early and late parts ofthe study, respectively; to Dr. Kenneth Walls, CDC, Atlanta, and Dr. Eisa Segura, Instituto Nacional de Diagnóstico e Investigación de la Enfermedad de Chagas, Buenos Aires, Argentina, for performing the serological tests; to Dr. Jorge Hernández, INDERENA, for confirming the identification of some of the animals captured; to Sharon Holstead, Mark Weitzel and Mathew Marshall, U.S. Peace Corps Volunteers, Olivia de Hincapié, Julio César Castro, Gerardo Cedeño and Jaime Torres, CIDEIM Assistants, and Genaro Useche, Heriberto Arias and Humberto Hurtado, for their technical assistance in the laboratory or for their collaboration in field activities; and to Dr. Robert J. Tonn, Parasite Diseases and Vector Control, PAHO/WHO, Washington, and Dr. Jorge E. Rabinovich, Instituto Venezolano de Investigaciones Científi-

ET AL.

cas, IVIC, Caracas, Venezuela, for reviewing the manuscript. REFERENCES

1. D'Alessandro, A, Barreto, P., and Duarte, C. A, 1971. Distribution of triatomine-transmitted trypanosomiasis in Colombia and new records of the bugs and infections. J. Med. Entomol., 8: 159-172. 2. D'Alessandro, A., Barreto, P., and Thomas, M., 1981. Nuevo registros de triatominos domiciliarios y extradomiciliarios en Colombia. Co10mb. Med., 12: 75-85. 3. Wells, E. A., D'Alessandro, A., Morales, G. A., and Angel, D., 1981. Mammalian wild1ife diseases as hazards to man and livestock in an area ofthe Llanos Orientales ofCo10mbia. 1. Wildl. Dis., 17: 153-162. 4. Tobie, E. J., von Brand, T., and Mehlman, B., 1950. Cultural and physiological observations on Trypanosoma rhodesiense and Trypanosoma gambiense. J. Parasitol., 36: 48-54. 5. D'Alessandro, A, 1972. New experimental vectors of Colombian Trypanosoma rangeli. J. Med. Entomol., 9: 187-195. 6. Whitlaw, J. T., Jr., and Chaniotis, B. N., 1978. Pa1m trees and Chagas' disease in Panama. Am. J. Trop. Med. Hyg., 27: 873-881. 7. Baker, J. R., Miles, M. A, Godfrey, D. G., and Barrett, T. V., 1978. Biochemica1 characterization of some species of Trypanosoma (Schizotrypanum) from bats (Microchiroptera). Am. J. Trop. Med. Hyg., 27: 483-491. 8. Zeledón, R., 1974. Epidemiology, modes oftransmission and reservoir hosts of Chagas' disease. Pages 51-85 in K. Ellicott, M. O'Connor, and G. E. W. Wo1stenho1me, eds., Trypanosomiasis and Leishmaniasis with Special ReJerence to Chagas' disease. CIBA Foundation Symposium 20 (N.S.), Associated Scientific Publishers, New York. 9. Ze1edón, R., andRabinovich,J. E., 1981. Chagas' disease: An ecologica1 appraisal with specia1 emphasis on its insect vectors. Annu. Rev. Entomol., 26: 101-133. 10. Gamboa, J., 1963. Comprobación de Rhodnius prolixus extradomici1iario en Venezuela. Bol. Ofic. Sanit. Panam., 54: 18-25. 11. Gamboa, J., 1973. La población silvestre de Rhodnius prolixus en Venezuela. Arch. Venezol. Med. Trop. Parasitol. Med., 5: 321-352. 12. Gamboa, J., 1974. Ecología dela tripanosomiasis americana (enfermedad de Chagas) en Venezuela. Bol. In! Dir. Malariol. San Am., 14: 3-20. 13. Tonn, R. J., Carcavallo, R. U., Ortega, R., and Carrasquero, B., 1976. Métodos de estudio de triatominos en el medio silvestre. Bol. Dir. Malariol. San. Amb., 16: 146-152. 14. Carcavallo, R. U., Tonn, R. J., Ortega, R., Betancourt, P., and Carrasquero, B., 1978. Notas sobre la biología, ecología y distribución geográfica de Rhodnius prolixus Sta!, 1859 (Hemiptera: Redu-

EPIDEMIOLOGY

15.

16.

17.

18.

19.

20. 21.

OF TRYPANOSOMA

viidae). Bol. Dir. Malario/. San. Amb., 18: 175198. Feliciangeli de Piñero, D., and Torrealba, J. W., 1977. Observaciones sobre Rhodnius prolixus (Hemiptera, Reduviidae) en su biotopo silvestre Copernicia tectorum. Bol. Dir. Malariol. San. Amb., 17: 198-205. Carcavallo, R., and Barreto, P., 1976. Una nueva especie de Rhodnius Sta1 (Hemiptera, Reduviidae, Triatominae) de Colombia. Bol. Dir. Malariol. San. Amb., 16: 176-184. Lent, H., and Wygodzinsky, P., 1979. Revision ofthe Triatominae (Hemiptera, Reduviidae), and their significance as vectors of Chagas' disease. Bul!. Am. Mus. Nat. Hist., 163: 125-520. Barreto, M., Barreto, P., and D'Alessandro, A., 1984. Psammolestes arthuri (Hemiptera: Reduviidae) and its parasite Telonomus capito (Hymenoptera: Scelionidae) in Colombia. J. Med. Entomol. (In press.) Tonn, R. J., Espinola, H., Mora, E., and Jiménez, J. E., 1978. Trampa de luz negra como método de captura nocturna de triatominos en Venezuela. Bol. Dir. Malariol. San. Amb., 18: 2530. Gómez-Nuñez, J. c., 1963. Notes sobre la ecología del Rhodnius prolixus. Bol. Dir. Malariol. San. Amb., 3: 330-335. Sherlock, 1. A., 1979. Vectores. Pages 42-88 in Z. Brener and Z. Andrade, eds., Trypanosoma cruzi e Doem;a de Chagas. Editora Guanabara Koogan, Rio de Janeiro.

CRUZI IN COLOMBIA

1095

22. D'Ascoli,A.,andGÓmez-Núñez,J.c., 1966. Notas sobre los medios de dispersion del Rhodnius prolixus StAl. Acta Ci. Venezo/., 17: 22-:25. 23. Gómez-Núñez, J. c., 1969. Resting places, dispersal and survival of C060 tagged adult Rhodnius prolixus. J. Med. Entomol., 6: 83-86. 24. Rabinovich, J. E., .Leal, J. A., and Piñero, D. F., 1980. Investigación de Campo sobre el Uso de Microhimenópteros para el Control Biológico de los Vectores del Trypanosoma cruzi. CONICIT, Caracas, Venezuela. 25. Lainson, J. J., Shaw, H. F., Fraiha, H., Miles, M. A., and Draper, C. c., 1979. Chagas' disease in the Amazon Basin. 1. Trypanosoma cruzi infections in sylvatic mammals, triatomine bugs and man in the State of Pará, north Brazil. Trans. R. Soco Trop. Med. Hyg., 73: 193-204. 26. Pinto Dias, J., and Borges Dias, R., 1979. Aspectos sociais da doen¡;a de Chagas. Rev. Goiana Med., 25: 157-168. 27. Forattini, O. P., Ferreira, O. A., da Rocha e Silva, E. O., and Rabello, E. X., 1978. Aspectos ecológicos da tripanossomiase Americana. XII. Varia