Mem Inst Oswaldo Cruz, Rio de
Janeiro, Vol. 97(1) 2002, pp. 53-59
Species Diversity and Flagellate Infections
in the Sand Fly Fauna near Porto Grande, State of Amapá, Brazil (Diptera:
Psychodidae. Kinetoplastida: Trypanosomatidae)
Rui A Freitas, Roberto D Naiff, Toby V Barrett+
Grupo de Biologia Vetorial e Eco-epidemiologia
de Trypanosomatidae na Amazônia, Coordenação de Pesquisas
em Ciências da Saúde, Instituto Nacional de Pesquisas da Amazônia,
Caixa Postal 478, 69011-970 Manaus, AM, Brasil
+Corresponding author. Fax: +55-92-643.3061. E-mail: tbarrett@inpa.gov.br
Financed by the Pilot Programme for the Protection
of Brazilian Rainforest/MCT - Subprograma C&T/PPD G-7
Received 4 May 2001
Accepted 16 August 2001
Code Number: oc02009
Forty-six species of Lutzomyia
and one species of Brumptomyia
were identified among 20,008 sand flies collected in central Amapá.
L. squamiventris maripaensis, L. infraspinosa, L. umbratilis and
L. ubiquitalis accounted for 66% of the specimens caught in light traps, and
L. umbratilis
was the commonest of the 16 species found on tree bases. Seven species of Lutzomyia
including L. umbratilis
were collected in a plantation of Caribbean pine. Sixty out of 511 female sand
flies dissected were positive for flagellates. Among the sand flies from which
Leishmania
was isolated, promastigotes were observed in the salivary glands and foregut
of 13 out of 21 females scored as having very heavy infections in the remainder
of the gut, reinforcing the idea that salivary gland invasion may be part of
the normal life cycle of Leishmania
in nature. Salivary gland infections were detected in specimens of L.
umbratilis, L.
whitmani and L.
spathotrichia. Parasites isolated from
L. umbratilis,
L. whitmani
and also from one specimen of L. dendrophyla
containing the remains of a bloodmeal, were compatible with Le.
guyanensis by morphology and behaviour
in hamsters.
Key words: Phlebotominae - diversity - distribution
- Leishmania - salivary glands - Amapá - Brazil
The phlebotomine sand fly fauna of the Brazilian
State of Amapá has been poorly studied in comparison with neighbouring
areas of the State of Pará and French Guyana (Brazil et al. 2000). Most
of the sand fly distribution maps in Young and Duncan (1994) are blank for this
part of Brazil, even though cutaneous leishmaniasis has long been recognized
as an important public health problem in the former Federal Territory (Lainson
& Shaw 1973). In this field report the faunistics of sand fly samples collected
mainly along the Perimetral Norte highway (BR-210) are analyzed. The eastern
end of this uncompleted highway runs west from the town of Porto Grande through
lowland rainforest in the lower Amapari basin. According to Backus et al. (1991)
this part of the highway lies in area 59 (Araguari) which they describe as the
nucleus of the Oyapock centre of endemism.
Although invasion of the vectors' salivary glands
by promastigotes is not currently thought to be a necessary stage in the transmission
of Leishmania by bite, the phenomenon has been reported in both naturally
infected (Arias & Freitas 1978, Naiff et al. 1991) and experimentally infected
sand flies (Killick-Kendrick et al. 1996). There have hitherto been so few of
such observations that some workers may be reluctant to accept that invasion
of the salivary glands is a normal part of the life cycle of Leishmania.
The presence of large numbers of infected Lutzomyia umbratilis
concentrated on the bases of trees in forest north of the Amazon River provides
an opportunity to re-examine this question. In the present report we show that
the prevalence of natural infection of Lutzomyia salivary glands by promastigotes,
presumably of Leishmania, may be much higher than was previously suspected.
MATERIALS AND
METHODS
Study areas - Most of the material was
collected in primary terra-firme lowland rainforest at km 7, km 17 and km 57
(from the town of Porto Grande) along the BR-210 highway. The site at km 7 is
1 km north of the highway at Recanto Ecológico Sonho Meu, where collections
were made along a tourist trail rising into the forest. At km 17 traps were
set in the forest on high ground near the edge of the highway. The site at km
57 is Assentamento Munguba, near 0043'N 051 53'W; about 5 km south of the
BR-210. This is a community of 28 households participating in a government sponsored
(Incra) settlement programme. The vegetation was still relatively undisturbed
apart from the access road itself and a clearing corresponding to the main villa.
Leishmania isolates IM-4677, IM-4678 and IM-4679 cultured from skin lesions
from three of the settlers are compatible with Le. guyanensis by morphology
and behaviour in hamsters.
The final site (BR-156: km 16) is 16 km southeast
of Porto Grande on the highway to Macapá, about 300 m from the edge of
the road. Light traps were set in a mature stand of Caribbean pine (Pinus
caribeae var. hondurensis) part of an extensive monocultural plantation
distant from any trace of the native forest.
Sampling - CDC miniature light traps were
suspended at approximately 1m from the ground. Tree-base samples were obtained
by aspiration of the trunk with a hand-held CDC light trap. Human bait catches
were carried out by four collectors between 18.00 h and 19.00 h. All specimens
collected on human bait were dissected. No attempt was made to select specimens
from tree-base samples randomly for dissection, and the material dissected is
probably biased in favour of fed and gravid females. Collectors: RAF, RDN and
Francisco Lima Santos (Inpa), 28 October-9 November 1999.
Diversity parameters - In the Tables,
alpha diversity is the Fisher-Williams index, which is a in the expression S
= a ln (1+N/a) where S
is the total number of species in the sample and N is the number of individuals.
In light trap samples abundance is expressed as the mean number of individuals
per trap per night. Dominance is the number of individuals of the most abundant
species divided by N.
Detection and isolation of flagellates -
Sand flies were placed in a drop of saline on a microscope slide and the head
separated from the thorax with a pair of needles, before drawing out the abdominal
and thoracic gut through the apex of the abdomen. In these preparations the
salivary glands and most of the foregut usually accompany the head, and the
diverticular crop of the foregut accompanies the midgut. The mid and hindgut
were examined first, and only if these presented heavy infections was the head
examined for flagellates. Positive guts were transferred to tubes of NNN blood-agar
culture medium. Unfortunately, these ran out just before the infected L.
spathotrichia was detected. Positive cultures were subsequently inoculated
intradermally in hamsters.
Material examined - Female sand flies
dissected for parasitological examination were preserved in gum-chloral on the
original microscope slides. Undissected specimens were cleared in sodium hydroxide
and phenol for identification and are at present stored in 70% ethanol in our
laboratory. Leishmanial isolates cryopreserved in liquid nitrogen will be deposited
in the Inpa culture collection (curator: Maricleide F Naiff, CPCS-Inpa).
RESULTS
Summaries of the sand fly samples are presented
in Tables I, II,
III, IV
and Figs 1-2.
Two males of the migonei species group, compatible with Fig. 66 of Young
and Duncan (1994) are provisionally associated with the females of the informally
named Lutzomyia sp. de Baduel (Floch & Abonnenc). L. squamiventris
maripaensis was the most abundant species in the light trap samples (26%)
and on human bait (84%). L. umbratilis was the most abundant species
on tree bases (87%) and the third most abundant species in the light trap samples
(14%). Species richness and diversity were unremarkable for the Amazon Region,
with a total of 46 species among 6,033 individuals captured in light traps.
Catches were high, particularly at km 7 where an average of 132 individuals
were captured per trap-night. Phlebotomines were very abundant on tree bases
at km 57, where 13,499 specimens including 11,964 L. umbratilis were
taken by a single collector in about an hour and a half. The 15 traps set for
one night in the Pinus plantation caught 11 sand flies of seven species,
including male and female L. umbratilis (Table
IV).
Of the 511 female sand flies dissected, 60 were
positive for flagellates (Table V). Of
the positive flies, 56 were L. umbratilis from tree bases, two were L.
dendrophyla from tree bases, one was L. whitmani from a CDC trap
at BR-210 km 17, and one was L. spathotrichia from a tree base at km
57. Of the 56 positive L. umbratilis, 55 were from km 57, equivalent
to an infection rate of 30% for tree-base L. umbratilis dissected at
this site.
Of the positive flies from which we did not succeed
in cultivating flagellates, the L. spathotrichia was infected with parasites
of similar appearance in vivo to those in L. umbratilis females from
which Leishmania was isolated. One L. umbratilis female without
a bloodmeal had a moderate infection in the anterior and posterior midgut, with
parasites tentatively identified in vivo as Trypanosoma sp.
Parasites were cultured from 40 Lutzomyia
females (Table VI). All of these isolates
contained Leishmania confirmed by hamster inoculation and are compatible
with Le. guyanensis by morphology and behaviour in hamsters. The primary
culture from L. dendrophyla also contained fusiform promastigotes thought
to be Endotrypanum sp. and which probably correspond to the parasites
observed in the Malpighian tubules (cf. Franco et al. 1997).
Salivary gland infections were recorded in 18
of the 60 positive flies dissected, including L. umbratilis, L. whitmani
and L. spathotrichia (Table V).
In flies from which Leishmania was isolated, salivary gland and foregut
infections were observed in 13 out of 21 females (62%) with heavy (++++) infections
of the abdominal and thoracic gut (Table VI).
DISCUSSION
The present survey supplements a list of 18 species
of Lutzomyia from lowland Amapá (Brazil et al. 2000) and
a preliminary report on an apparently more diverse phlebotomine fauna from the
Serra do Navio area (Souza et al. 2001). Many of the species listed here were
previously known from French Guyana and/or Pará in Brazil, but our records
represent significant extensions of the known distribution particulary for L.
evangelistai, L. damascenoi, L. ininii (first record for Brazil),
L. williamsi, L. dreisbachi and L. chassigneti. The records
of L. squamiventris maripaensis and L. yuilli pajoti help fill
the gap between the known northern and southern localities of these taxa. Species
common in collections from Amazonas and western Pará and conspicuous
by their absence from the present samples include L. olmeca nociva, L.
chagasi and L. carrerai. These species are also absent from the known
fauna of French Guyana.
The recently described Lutzomyia campograndensis
Oliveira, Andrade Filho, Falcão & Brazil 2001 is known to occur
in Amapá (Oliveira et al. 2001) and may be represented among the four
specimens we have identified as L. lutziana. These two species are so
similar morphologically that at present we are unable to separate them confidently
on the basis of the published descriptions alone.
L. umbratilis, a vector of Le.
guyanensis, was previously thought to be incapable of adapting to pine
plantations (Lainson 1988). Our finding of L. umbratilis and other species
in this anthropic environment may be related to the age of the plantation and
the presence of deer and other mammals feeding on the seeds of the mature trees.
Le. guyanensis has previously been
isolated in the Jari area of Pará, approximately 180 km SW of the present
study area, from L. umbratilis, L. whitmani and L. anduzei
by Ryan et al. (1987). One out of six L. spathotrichia in that collection
was also infected with flagellates, but these were apparently not isolated or
identified. L. spathotrichia is classified in the subgenus Lutzomyia,
not known for including vectors of Le. guyanensis, but L. gomezi
in the same cruciata Series is suspected as a vector of the closely related
Le. panamensis (see Christensen et al. 1983). Our finding of L.
spathotrichia with promastigotes in the mouthparts and salivary glands suggests
that further attention to this anthropophilic (Young & Duncan 1994) and
locally abundant (Table II) species as
a possible vector of Le. guyanensis could be rewarding.
The isolate IM-4685 from L. dendrophyla
has now been confirmed as Le. guyanensis by enzyme electrophoresis (AR
Franco, unpublished observations). The leishmanial parasites may however have
been confined to the remains of the bloodmeal in the anterior midgut, and we
do not suggest that L. dendrophyla is likely to be a natural host of
this parasite.
In Table VI,
the combinations of absent or fresh bloodmeals with medium development of the
ovarioles observed in L. umbratilis suggest that gonadotrophic concordance
may not always be simple (one bloodmeal per oviposition) in this species, and
that at least some of these females required a second bloodmeal for egg maturation.
Ingestion of a second bloodmeal has been shown to influence the distribution
of promastigotes of Le. amazonensis and Le. mexicana in L.
evansi under laboratory conditions (Vivenes et al. 2001).
Evidence exists for at least two mechanisms for
the transmission of Leishmania by bite in the absence of invasion of
the vectors' salivary glands (Killick-Kendrick 1979, Schlein et al. 1992). However,
Killick-Kendrick et al. (1996) obtained experimental salivary gland infections
in Phlebotomus duboscqi that had been allowed to feed on a suspension
of Le. (Le.) tropica amastigotes. Promastigotes in the
salivary glands of wild sand flies had previously been reported for L. anduzei
and L. umbratilis infected with Le. guyanensis and in a female
of the L. shannoni species group (subgenus Psathyromyia) infected
with uncharacterised parasites (Arias & Freitas 1978) as well as in L.
sq. squamiventris infected with Le. (V.) naiffi (see
Naiff et al. 1991).
Failure to detect salivary gland infections in
wild sand flies is only meaningful in relation to the number of glands examined
and the overall infection rates in the samples of sand flies dissected; and
the scarcity of positive reports may be due in part to the difficulty of finding
naturally infected females in many Leishmania/vector systems. As all
of the reports of natural salivary gland infections are from the same group
at Inpa, Killick-Kendrick et al. (1996) were understandably cautious when concluding
that invasion of the salivary glands could, conceivably, be a normal part of
the life cycle of Leishmania. Independent confirmation of the phenomenon,
if possible supported by electron microscopy of thin sections of naturally infected
flies, would therefore be of great interest. In Brazil, abundant material for
study can be found on tree bases in terra-firme rainforest north of the Amazon
River, from Amapá to Manaus.
ACKNOWLEDGEMENTS
To the Fundação Nacional de Saúde,
Macapá, particulary Dr Paulo Almeida Chavier, Director; Mr Raimundo Nonato
de Anjos Freire, Chief Entomologist; Mr José Cláudio C Mendes
and Mr Luiz Otávio C Nascimento, Technicians; and to the Fundação
Nacional de Saúde, Unidade Mista de Porto Grande, particulary Dr José
Maria Cordeiro da Silva, Director; Dr Arlene de Paiva Brandão, Chief
Nurse, and Nadir Ferreira Lamião, Technician, for outstanding logistic
support in Amapá.
REFERENCES