Blood-feeding performance of nymphs and adults of Triatomabrasiliensis on human hosts

Alessandra A. Guarneri a,1, Lileia Diotaiuti a, Nelder F. Gontijo b, AlbertoF. Gontijo c, Marcos H. Pereira b,*

a Centro de Pesquisas Rene Rachou, Fiocruz, Caixa Postal 1743, 30190-002, Belo Horizonte, MG, Brazilb Departamento de Parasitologia do Instituto de Ciencias Biologicas da Universidade Federal de Minas Gerais, Caixa Postal 486, 31270-

901, Belo Horizonte, MG, Brazilc Colegio Tecnico da Universidade Federal de Minas Gerais, Belo Horizonte, MG, Av. Antonio Carlos, 6627, 31270-901, Belo Horizonte,

MG, Brazil

Received 18 June 2002; received in revised form 23 March 2003; accepted 1 April 2003

Abstract

The blood-feeding behaviour of nymphs and adults of Triatoma brasiliensis fed on the forearm of human volunteers

was studied by electronic monitoring of the cibarial pump. Parameters of total contact time (TT), initial weight (IW),

weight gain (WG), ingestion rate (IR), pump frequency (F), quantity of liquid ingested per cibarial pump stroke (QLC)

and non-ingestive time (NIT) (cumulative probing time and pumping interruptions during blood feeding) were

measured. Protein profile (SDS�/PAGE) and quantity of proteins of salivary gland extracts (QP) were also determined

for each stage. The TT reflects the feeding performance of the insects and differed between instars, varying between

18.39/2.5 min for the first instar and 33.99/2.3 min for the fifth instar. The observed increase in the IR when comparing

different instars was related to the increase in the cibarial pump volume inferred from the QLC data. During

development, the volume of the cibarial pump grew asymmetrically determining the different contact times observed

among the instars. Males and females presented a remarkable sexual dimorphism in respect to the volume of the cibarial

pump, females showing a better performance compared to males. Despite the differences, the results show that each of

the development stages of T. brasiliensis was able to obtain a relatively fast bloodmeal, with few interruptions and

without causing pain, providing further evidence of the capacity of this species to adapt to domestic environments.

# 2003 Elsevier Science B.V. All rights reserved.

Keywords: Feeding behaviour; Triatoma brasiliensis ; Interaction triatomine�/host; Blood-feeding performance; Human host

1. Introduction

Triatoma brasiliensis (Hemiptera, Reduviidae,

Triatominae) is now considered the most impor-

tant vector of Trypanosoma cruzi in Brazil, where

it is widely distributed in the nine states of the

* Corresponding author. Tel.: �/55-31-499-2867; fax: �/55-

31-499-2970.

E-mail address: marcoshp@mono.icb.ufmg.br (A.A.

Guarneri).1 Part of a thesis was submitted to the Instituto Oswaldo

Cruz, FIOCRUZ, Rio de Janeiro, Brazil, in partial fulfilment of

the requirements for a doctoral degree.

Acta Tropica 87 (2003) 361�/370

www.elsevier.com/locate/actatropica

0001-706X/03/$ - see front matter # 2003 Elsevier Science B.V. All rights reserved.

doi:10.1016/S0001-706X(03)00121-9

Northeast region (Silveira et al., 1984). Thisspecies is able to colonise houses and peridomici-

lary areas, and is also widely distributed in sylvatic

habitats where it is mainly found amongst rock

piles associated with various species of vertebrates.

This increases its importance as a vector, since it

can re-colonise domestic habitats from which it

was eliminated through insecticide spraying (Alen-

car, 1987; Diotaiuti et al., 2000).T. brasiliensis as well as other Triatominae are

obligatory haematophagous insects. They are

vessel feeders, removing blood directly from the

venules and arterioles. The events that make up

the feeding process include detection of the host,

active exploratory movements of the mouthparts

on the surface of the skin, biting or penetration

(insertion of the mouthparts into the skin), loca-lisation of blood within a blood vessel, ingestion of

blood and then cessation of feeding (Friend and

Smith, 1977). Duration of contact is an important

factor in the interaction between haematophagous

arthropods and their hosts, since defensive beha-

viour of the hosts can reduce feeding success.

Experiments with Triatominae have shown that

reduction of bloodmeal size associated with in-creases in bug density is modulated by host

perception of the bite (Schofield et al., 1986;

Schofield, 1994) and determined by decrease in

contact time (Pereira et al., 1995, 1998). Thus,

species that possess higher total ingestion rates

tend to achieve higher nutritional status, and reach

higher population densities, as has been observed

for Triatoma infestans and Rhodnius prolixus , themost important vectors of T. cruzi in South and

Central America, respectively (Sant’Anna et al.,

2001).

Parameters of feeding behaviour are influenced

by intrinsic characteristics of the feeding apparatus

(such as the size of the cibarial pump, which varies

between species), and by host factors related to

haemostasis, such as probing time, frequency ofcontractions of the cibarial pump and interrup-

tions during the engorgement phase. Thus, the

characteristics of the insect and the influence of

host haemostasis determine the duration of con-

tact during the blood feeding. Among the Tria-

toma species so far studied, total contact time of

fifth instar nymphs with their hosts ranged from

36.29/4.7 min for Triatoma pseudomaculata fed onmice to 9.99/1.1 min for T. infestans fed on

pigeons (Guarneri et al., 2000). Studies on triato-

mine biology have shown different contact times

among the nymphal instars when insects of the

same species were fed on the same host (Zeledon et

al., 1977; Goncalves et al., 1988, 1997; Soares et

al., 2000). Here, we evaluated the parameters of

feeding behaviour that influence the contact timeof the developmental stages of T. brasiliensis fed

on human hosts.

2. Materials and methods

2.1. The insects

T. brasiliensis used in this study were the second

generation from parents originally collected from

human dwellings in Novo Oriente, Ceara state,

Brazil. They were reared in the laboratory under

semi-controlled conditions (289/2 8C and 659/

10% RH) and fed weekly on mice or chickens.

Insects with starvation periods of 4 (first instar

nymphs), 7 (second, third and fourth instar

nymphs) and 14 (fifth instar nymphs and adults)

days after eclosion or ecdysis were used in the

experiments.

2.2. Feeding behaviour

The technique used to study feeding behaviour

was modified from that used by Guarneri et al.(2000). An electrode comprising a short length of

gold wire (diameter 89.5 mm) was fixed to the

forearm of a human volunteer (the volunteers were

three of the authors). A second electrode was

connected to a metal mesh, which was fixed to a

container used as substrate for the insects. These

two electrodes detected an electrical signal from

the cibarial pump, which was filtered to rejectfrequencies above 17 Hz (with a slope of �/45 dB/

octave), amplified 210 times and collected via a

digital data acquisition plate (ADC100†-Pico

Technology Limited, UK) connected to an IBM-

AT compatible microcomputer. The electrode and

the metal mesh were moistened with an electrolytic

A.A. Guarneri et al. / Acta Tropica 87 (2003) 361�/370362

gel (Regisgraf-Gel†) to facilitate passage of elec-trical signals.

The insects were weighed immediately before

(initial weight, IW) and after feeding to determine

weight gain (WG). Each insect was allowed to feed

until it removed its mouthparts without trying to

probe again within 1 minute. The engorgement

phase was defined as beginning after at least ten

consecutive characteristic peak-like signals attrib-uted to the cibarial pump contractions. All insects

that at least doubled their initial weights were

included in the analysis. Ten insects were used in

each group. Care was taken to minimise light,

odour, noise, vibration and handling of the insects

before testing.

From the cibarial pump records, the total

contact time (TT) was defined as the time duringwhich the insect mouthparts remained inserted

into the host skin. Non-ingestive time (NIT)

(cumulative probing time (PT) and interruption

time (IT)) was also determined. Cumulative prob-

ing time was defined as the time from initial

insertion of the mouthparts into the host up to

the beginning of the engorgement phase. If the bug

ended an initial probing and restarted another oneelsewhere, the first probing time was added to the

second and so on. Interruption time was defined as

the time during which the insects were not pump-

ing blood after the engorgement phase had begun.

The quantity of liquid ingested per cibarial pump

contraction (QLC, mg) was obtained by dividing

WG (mg) by the total number of cibarial pump

contractions during engorgement. Pump frequency(F) represents the total number of cibarial pump

contractions divided by the functioning time of the

cibarial pump (time (s) in which the pump was

effectively working, without probing or interrup-

tions). Ingestion rate (IR, mg/min) was calculated

by multiplying QLC by F.

Statistical analysis used Statistica-version 5.0,

Statsoft Inc., 1989�/1993. The Kolmogorov�/Smir-nov test was used to determine normality for each

variable, and those that followed a normal dis-

tribution were compared by Student t-test or

analysis of variance (ANOVA). In the case of

ANOVA tests, post hoc comparisons were made

using the Scheffe test. The Mann�/Whitney and

Kruskal�/Wallis non-parametric tests were used

for variables with a non-normal distribution. Theaccepted significance level was p B/0.05 (95%).

2.3. Saliva analysis

Pairs of salivary glands (D1 and D2 lobes) from

nymphs of each instar (n�/5; K�/3) were dissected

into 0.5 ml conical plastic tubes containing phos-

phate-buffered saline (150 mM NaCl and 10 mMsodium phosphate, pH 7.2). Whole glands were

frozen at �/20 8C until used 1�/10 days later.

Glands were then disrupted in presence of protease

inhibitor (Pepstatin A, 5 mg/ml), centrifuged at

9000�/g for 5 min and the supernatant removed

for use in the assays. The quantity of proteins per

pair of salivary glands (QP) was determined using

bovine serum albumin as a standard (Bradford,1976). Electrophoresis (SDS�/PAGE) of salivary

gland extracts was carried out on a 12% poly-

acrylamide slab gel (Hubbard and Lazarides,

1979) and protein bands were visualised by silver

nitrate staining (Tunon and Johansson, 1984). For

the analysis of electrophoretic patterns, a matrix

based on the presence:absence of bands was

constructed. A phenogram based on shared bandswas constructed using the Dice similarity coeffi-

cient (Dice, 1945) to produce a matrix of similar-

ity, which was then used for unweighted pair

group method analysis (UPGMA).

3. Results

The feeding behaviour of each of the fivenymphal stages, males and females of T. brasilien-

sis was analysed by means of electronic monitor-

ing of the cibarial pump (Table 1). On average the

insects probed only once, with the exception of

first instar nymphs which bit a mean number of

1.79/0.2 times. Duration of the cumulative prob-

ing times on human hosts was variable (Fig. 1) but

no significant differences were observed betweenthe developmental stages (Kruskal�/Wallis, n.s.).

Irrespective of their stage of development, 50% of

the insects located a blood vessel within 55 s.

Probing times greater than 120 s were only

observed for first instar nymphs and adults.

None of the volunteers (AAG, NFG, MHP)

A.A. Guarneri et al. / Acta Tropica 87 (2003) 361�/370 363

reported feeling pain during the feeding process of

the insects.

Once the cibarial pump began to function, few

interruptions were observed in the groups studied.

Most of the interruptions occurred near the end of

the meal, before the insect had removed its

mouthparts. None of the interruptions lasted

more than 2 min.

The mean frequency of contractions of the

cibarial pump differed among developmental

stages (Table 1; ANOVA; p B/0.05) the highest

mean value being observed for second instar

nymphs (3.39/0.2 contractions/s) and the lowest

for those of the fourth instar (2.49/0.2 contrac-

tions/s) (Table 1). In addition, the frequency of

contractions of the cibarial pump varied during

the engorgement phase in all stages (Fig. 2). In

about 80% of the insects, lowest frequencies were

observed at the initiation of the period, the values

increasing and being maintained until nearly 85%

of the meal had been ingested (Scheffe, p B/0.05).

After this period, a reduction occurred in the

frequency of contractions for all the developmen-

tal stages, with the exception of the fourth instar

(Scheffe, p B/0.05).The exponential increase in body weight of the

insects during development (almost 84 times from

first instar to adult) produced a high correlation

between initial weight and certain parameters that

also increased exponentially, such as WG (r2�/

0.89), QLC (r2�/0.94) and QP (r2�/0.99). How-

ever, when these parameters were normalised with

the initial weight, the observed increase of the bugs

during their development was not proportional

(Fig. 3a�/c) (normalisation refers to how much a

determined parameter increased when compared

with the respective IW). The normalised WG was

similar among the instar stages (ANOVA; n.s.).

However, in the adults this parameter was less

than the mean value obtained for the nymphal

stages (Fig. 3a; ANOVA; p B/0.05). Normalised

QLC differed among the developmental stages

(Fig. 3b; ANOVA; p B/0.05), this parameter tend-

ing to decrease during the development of the

insect and reaching the lowest mean values

amongst males. Normalised QP was proportional

in all the developmental stages except in the fifthTa

ble

1

Pa

ram

eter

so

ffe

edin

gb

ehavio

ur

of

T.

bra

sili

ensi

sfe

do

nh

um

an

ho

sts

Dev

elo

pm

enta

lst

ag

e1

stin

star

2n

din

star

3rd

inst

ar

4th

inst

ar

5th

inst

ar

Fem

ale

sM

ale

s

Cu

mu

lati

ve

pro

bin

gti

me

(s)

(med

ian

)69.7

31.0

36.7

40.3

41.8

59.2

60.8

Init

ial

wei

gh

t(m

g)

(mea

n9

/S.E

.)1

.49

/0.0

54

.39

/0.3

10

.69

/0.9

18

.99

/1.8

68

.09

/4.

05

11

6.29

/11

.21

17

.59

/9.5

Wei

gh

tg

ain

(mg

)(m

ean9

/S.E

.)6

.59

/0.3

24

.09

/1.3

46

.39

/4.9

85

.99

/13

.33

59

.39

/24

.42

28

.19

/17

.91

45

.39

/8.3

Pu

mp

freq

uen

cy(c

on

tra

ctio

ns/

seg

)(m

ean9

/S.E

.)2

.89

/0.2

3.39

/0.2

2.69

/0.2

2.49

/0.2

3.09

/0.2

2.59

/0.2

2.59

/0.2

Qu

an

tity

of

blo

od

inges

ted

per

cib

ari

al

pu

mp

con

tract

ion

(mg

)(m

ean9

/S.E

.)2

.79

/0.1

7.19

/0.3

12

.69

/1.0

26

.99

/2.1

62

.89

/3.4

88

.99

/5.9

67

.99

/4.9

Ing

esti

on

rate

(mg

/min

)(m

ean9

/S.E

.)0

.49

/0.0

31

.49

/0.1

2.09

/0.3

3.89

/0.4

11

.39

/0.9

13

.69

/1.6

10

.49

/1.2

To

tal

con

tact

tim

e(m

in)

(mea

n9

/S.E

.)1

8.39

/2.5

18

.79

/2.1

25

.59

/3.1

23

.59

/3.2

33

.99

/2.3

21

.19

/2.5

17

.29

/1.6

Pro

tein

sp

erp

air

of

sali

va

ryg

lan

ds

(mg

)(m

ean9

/S.E

.)0

.49

/0.0

31

.29

/0.2

3.39

/0.1

5.19

/0.5

11

.19

/0.6

35

.49

/3.9

30

.39

/2.3

A.A. Guarneri et al. / Acta Tropica 87 (2003) 361�/370364

instar, for which the lowest relative values were

observed (Fig. 3c).

Despite the different F found amongst instars,

the observed increase on IR (calculated from F

and QLC) during the development of the insects

was due mainly to the increase in the volume of the

cibarial pump (Table 1).

The mean contact time was different between

the nymphal stages, varying between 18.39/2.5

min for the first instar and 33.99/2.3 min for the

fifth instar (Table 1; Anova, p B/0.05). Fig. 4

shows the normalised QLC in each instar related

to their respective contact time. The asymmetric

increase of the QLC during the development of

nymphs was inversely correlated with the different

contact times, such that stages with higher normal-

ised QLC were those that presented the lowest

bloodmeal duration (r2�/0.98). When the adult

Fig. 1. Rank order of cumulative probing time for T. brasiliensis fed on human hosts (each symbol represents one insect).

Fig. 2. Frequency of contractions of the cibarial pump during the ingestion period of T. brasiliensis fed on human hosts.

A.A. Guarneri et al. / Acta Tropica 87 (2003) 361�/370 365

stage was taken into consideration, the contacttime of males and females was similar (Table 1; t -

test, n.s.). However, despite having similar initial

weights, the females presented higher values for

weight gain and quantity of liquid ingested per

contraction (Table 1; t-test, p B/0.05).

In order to visualise the electrophoretic profile

of the proteins present in salivary glands extracts

of T. brasiliensis , 2.5 mg of proteins from each ofthe developmental stages were analysed (Fig. 5).

The electrophoretic profiles differed between de-

velopmental stages, although bands of 77.4, 59.6,

54.3, 33.8, 29.4 and 19.4 kDa (Fig. 5, arrowed)

were common to all of them. A total of 55 bands

was selected for comparative analysis. The matrix

of similarity was calculated in accordance with the

degree of band sharing between each pool ofsalivary gland extracts (Dice, 1945), from which

the unweighted pair-group method was generated

(Fig. 6). The tree shows that the earlier stages of T.

brasiliensis differed from the others with a simi-

larity coefficient of 0.73. As the development

progresses, the similarity of salivary glands profiles

increases, being the males and females equivalent.

4. Discussion

Modifications of the technique to monitor

electrical activity of the cibarial pump (Guarneri

et al., 2000) allowed the feeding behaviour of

triatomines to be studied during the initial stages

of development, since contact involved the legs

instead of electrodes fixed to the thorax of theinsect under test. These modifications also allow

the technique to be applied to other small species

of insects as has recently been demonstrated in a

study with ants (Josens and Pereira, 2002). The

normalised weight gain obtained for the different

developmental stages in the present study (Fig. 3a)

are similar to those obtained by Szumlewicz (1976)

in her study of the biology of T. brasiliensis in thelaboratory, suggesting that the modified technique

described here did not interfere with the capacity

of the insects to obtain a bloodmeal.

Feeding behaviour studies have shown that the

parameters which can influence triatomine�/host

contact time are the pump contraction rate,

Fig. 3. The ratio between weight gain (WG, a), quantity of

liquid ingested per contraction (QLC, b), quantity of salivary

proteins per insect (QP, c) and initial weight (IW) of the

different developmental stages of T. brasiliensis fed on human

hosts.

A.A. Guarneri et al. / Acta Tropica 87 (2003) 361�/370366

quantity of liquid ingested per pump contraction

and non-ingestive periods (probing plus interrup-

tions) (Guarneri et al., 2000; Sant’Anna et al.,

2001). Based on this, contact time can be described

by TC�/WG/(F.QLC)�/NIT (TC, min; WG, mg;

F, contractions/min; QLC, mg; NIT, min).

In spite of representing only 6% of the contact

time, probing time is a critical period in the feeding

process of triatomines. Insertion of the mouth-

parts and the movements of the maxillae in the

host skin both lead to tissue damage, which is

responsible for liberation of mediators and exposi-

tion of molecules that will trigger haemostasis and

Fig. 4. Contact time (TT) as a function of the normalised quantity of liquid ingested per cibarial pump contraction (QLC/IW) of

nymphs of T. brasiliensis fed on human hosts.

Fig. 5. Electrophoretic profile (SDS�/PAGE) of salivary gland extracts of T. brasiliensis. Lanes: (1) molecular mass markers (Kda); (2)

first instar nymph; (3) second instar nymph; (4) third instar nymph; (5) fourth instar nymph; (6) fifth instar nymph; (7) female; (8) male.

Arrows indicate some bands present in all stages. Each lane contains the salivary gland extracts of a pool of 13 insects.

A.A. Guarneri et al. / Acta Tropica 87 (2003) 361�/370 367

inflammation reactions (Ribeiro, 1995). In addi-

tion, these mechanical and physiological events

can also induce the generation of impulses at nerve

terminations near to the site of biting. In fact,

work with human volunteers suggests that the

period of greatest perception corresponds to the

insertion of the mouthparts into the host skin

(Schofield et al., 1986). In this sense, the prolonged

probing of the first instar nymphs and adults could

be influencing the capacity of these stages to

obtain a bloodmeal in non-restrict hosts.

The frequency of cibarial pump contractions

varied during the course of feeding, diminishing in

the final 15% of the meal. A decline in the

contraction frequency at the end of the feeding

process was also observed for R. prolixus fed on

artificial diets, which was deduced to reflect an

increase in the time taken to empty the pump as

back-pressure due to stretching of the abdominal

cuticle increased (Smith, 1979). If so, the fact that

males showed a lower weight gain but a reduction

in the contraction frequency at the end of feeding

similar to that of the females, suggests that males

are less able to expand their abdomens during

feeding.

Furthermore, comparison between the mean

frequency of contractions of the cibarial pump of

fifth instar nymphs of T. brasiliensis fed on

humans (3.09/0.2) with similar data obtained by

Guarneri et al. (2000) for nymphs fed on pigeons

(3.99/0.2) and mice (3.09/0.2) provides further

evidence to suggest that feeding on mammals is

more difficult for triatomines than taking blood

from birds. The fewer interruptions observed

during feeding on human hosts, when compared

with other hosts, make this species fairly efficient

to enter and settle inside the houses.

The electrophoretic profiles of the salivary gland

extracts were variable although some bands were

present in all the stages. The tree constructed from

salivary glands extracts profiles showed a gradual

change occurring among the different stages of T.

brasiliensis . These alterations could be a strategy

of the species to avoid immunogenic reactions

Fig. 6. Unweighted pair-group method of T. brasiliensis based on pairwise shared salivary proteins bands. The numbers on the

horizontal scale were derived from the Dice similarity coefficient. The codes on the right indicate the development stage: N1, first instar

nymph; N2, second instar nymph; N3, third instar nymph; N4, fourth instar nymph; N5, fifth instar nymph; F, female; M, male. The

vertical bar represents the average level of similarity between all the pair samples analysed (phenon line).

A.A. Guarneri et al. / Acta Tropica 87 (2003) 361�/370368

from the host. Saliva is continuously liberated

during the feeding process, being necessary for the

process of finding blood vessels during probing

and keeping flow of blood through feeding canal.

In this sense, salivary function may influence

probing time (Ribeiro, 1995) and also the modula-

tion of pump frequency on live hosts (Sant’Anna

et al., 2001). Although the salivary profiles of each

stage were distinctive, it was not possible yet to

correlate this variability with the differences on

feeding performance of the insects. Considering

that NIT was reduced (approximately 8% of TT)

and normalised WG was similar among the

nymphal stages, the parameters which influenced

the duration of the bloodmeal in these groups were

those that affected with the ingestion rate: pump

frequency and quantity of liquid ingested per

pump contraction. Although both parameters

differed between the instars, a higher correlation

between IR and QLC was found, showing that the

last parameter was that mainly responsible for the

increase of IR. As cibarial pump size inc reased

asymmetrically during the development of

nymphs, it is possible to suggest that this factor

was responsible for the different contact times

when the nymphs were fed on the same host. This

is different to observe when nymphs of the same

developmental stage are fed on different hosts, in

which the variations of contact time are related

mainly to the different pump frequency and

interruptions (Guarneri et al., 2000, Sant’Anna et

al., 2001).

A proportionally higher QLC would allow the

insect to have a better performance, obtaining a

proportionally larger bloodmeal in a similar per-

iod of contact with the host. The same rationale

could be applied to the differences encountered

between males and females. A proportionally

higher QLC of the females allows them to show

a better performance compared to males, which

may be an adaptation to increase their fitness in

situations of nutritional stress. It is possible that

the reduced size of insects observed in situations of

high density (Schofield, 1980; Dujardin et al.,

1999) reflects selection for individuals with a high

QLC/IW ratio, which are thus more able to feed

within a shorter period.

The insects used in the present study weredescended from triatomines collected inside

houses. Although comparative studies with other

species of triatomines on human hosts are lacking,

our results show that despite the differences, each

of the development stages of T. brasiliensis was

able to obtain a relatively fast bloodmeal, with few

interruptions and without causing pain, providing

further evidence of the capacity of this species toadapt to domestic environments.

Acknowledgements

We would like to thank Dr. C.J. Schofield for

helpful comments and for reviewing the manu-

script and to S.E. Barbosa for help on EGS

analysis. This study was supported by Fundacao

de Amparo a Pesquisa do Estado de Minas Gerais

(FAPEMIG), Conselho Nacional de Desenvolvi-mento Cientıfico e Tecnologico (CNPq) and Cen-

tro de Pesquisas Rene Rachou/Fundacao Oswaldo

Cruz (CPqRR/FIOCRUZ), and benefited from

international collaboration through the ECLAT

network.

References

Alencar, J.E., 1987. Historia natural da doenca de Chagas no

Estado do Ceara. Imprensa Universitaria da UFC, For-

taleza, 339 pp.

Bradford, M.M., 1976. A rapid and sensitive method for the

quantitation of microgram quantities of protein utilising the

principle of protein�/dye binding. Anal. Biochem. 72, 248�/

254.

Dice, L.R., 1945. Measures of the amount of ecological

association between species. Ecology 26, 297�/302.

Diotaiuti, L., Faria Filho, O.F., Carneiro, F.C.F., Dias, J.C.P.,

Pires, H.H.R., Schofield, C.J., 2000. Operational aspects of

Triatoma brasiliensis control. Cad. Saude Publica 16 (Suppl.

2), 61�/67.

Dujardin, J.P., Steindel, M., Chavez, T., Machane, M., Scho-

field, C.J., 1999. Changes in the sexual dimorphism of

Triatominae in the transition from natural to artificial

habitats. Mem. Inst. Oswaldo Cruz 94, 565�/569.

Friend, W.G., Smith, J.J.B., 1977. Factors affecting feeding by

blood-sucking insects. Annu. Rev. Entomol. 22, 309�/331.

Goncalves, T.C.M., Victorio, V.M.N., Jurberg, J., Cunha, V.,

1988. Biologia do Triatoma vitticeps (Stal, 1859) em

condicoes de laboratorio (Hemiptera: Reduviidae: Triato-

A.A. Guarneri et al. / Acta Tropica 87 (2003) 361�/370 369

minae). I. Ciclo evolutivo. Mem. Inst. Oswaldo Cruz 83,

519�/523.

Goncalves, T.C.M., Cunha, V., Oliveira, E., Jurberg, J., 1997.

Alguns aspectos da biologia de Triatoma pseudomaculata

Correa & Espınola, 1964, em condicoes de laboratorio

(Hemiptera: Reduviidae: Triatominae). Mem. Inst. Oswaldo

Cruz 92, 275�/280.

Guarneri, A.A., Diotaiuti, L., Gontijo, N.F., Gontijo, A.F.,

Pereira, M.H., 2000. Comparison of feeding behaviour of

Triatoma infestans , Triatoma brasiliensis and Triatoma

pseudomaculata in different hosts by electronic monitoring

of the cibarial pump. J. Insect Physiol. 46, 1121�/1127.

Hubbard, B.D., Lazarides, E., 1979. The co-purification af

actin and desmin from chicken smooth muscle and their co-

polymerization in vitro to intermediate filaments. J. Cell

Biol. 80, 166�/182.

Josens, R.B., Pereira, M.H., 2002. Cibarial pump activity

during nectar feeding in the ant Camponotus mus . 17a.

Reuniao Anual das Federacoes de Sociedades de Biologia

Experimental, Salvador, Bahia.

Pereira, M.H., Penido, C.M., Martins, M.S., Diotaiuti, L.,

1995. Triatoma infestans is more efficient than Panstrongy-

lus megistus in obtaining blood meals on non anaesthetised

mice. Mem. Inst. Oswaldo Cruz 90, 765�/767.

Pereira, M.H., Penido, C.M., Martins, M.S., Diotaiuti, L.,

1998. Comparative kinetics of bloodmeal intake by Tria-

toma infestans and Rhodnius prolixus , principal vectors of

Chagas disease. Med. Vet. Entomol. 12, 84�/88.

Ribeiro, J.M.C., 1995. Blood feeding arthropods*/live syringes

or invertebrate pharmacologists. Infect. Agent Dis. 4, 143�/

152.

Sant’Anna, M.R.V., Diotaiuti, L., Gontijo, A.F., Gontijo,

N.F., Pereira, M.H., 2001. Feeding behaviour of morpho-

logically similar Rhodnius species: influence of mechanical

characteristics and salivary function. J. Insect Physiol. 47,

1459�/1465.

Schofield, C.J., 1980. Density regulation of domestic popula-

tions of Triatoma infestans in Brazil. Trans. R. Soc. Trop.

Med. Hyg. 74, 761�/769.

Schofield, C.J., 1994. Triatominae: Biologia y Control. Euro-

communica Publications, West Sussex, pp. 20�/29.

Schofield, C.J., Williams, N.G., Marshall, T.F.C., 1986. Den-

sity-dependent perception of triatomine bug bites. Ann.

Trop. Med. Parasitol. 80, 351�/358.

Silveira, A.C., Feitosa, V.R., Borges, R., 1984. Distribuicao de

triatomıneos capturados no ambiente domiciliar, no perıodo

de 1975/83, Brasil. Rev. Bras. Malariol. Doencas Trop. 36,

15�/312.

Smith, J.J.B., 1979. Effect of diet viscosity on the operation of

the pharyngeal pump in the blood-feeding bug Rhodnius

prolixus . J. Exp. Biol. 82, 93�/104.

Soares, R.P.P., Evangelista, L.G., Laranja, L.S., Diotaiuti, L.,

2000. Population dynamics and feeding behaviour of

Triatoma brasiliensis and Triatoma pseudomaculata , main

vectors of Chagas disease in Northeastern Brazil. Mem.

Inst. Oswaldo Cruz 95, 151�/155.

Szumlewicz, A.P., 1976. Laboratory colonies of Triatominae,

biology and population dynamics. In: New Approaches in

American Trypanosomiasis Research�/PAHO. Scientific

Publication, 318, Brockton, UK, pp. 63�/82.

Tunon, P., Johansson, K.E., 1984. Yet another improved silver

staining method for the detection of proteins in polyacry-

lamide gels. J. Biochem. Biophys. Methods 9, 171�/179.

Zeledon, R., Alvarado, R., Jiron, L.F., 1977. Observations on

the feeding and defecation patterns of three triatomine

species (Hemiptera: Reduviidae). Acta Trop. 34, 65�/77.

A.A. Guarneri et al. / Acta Tropica 87 (2003) 361�/370370