BIOLOGY AND COLONIZATION OF THE SAND FLY
Lutzomyia diabolica (HALL) (DIPTERA: PSYCHODIDAE)
WITH NOTES ON ITS POTENTIAL RELATIONSHIP TO HUMAN
CUTANEOUS LEISHMANIASIS IN TEXAS, USA
PHILLIP G. LAWYER
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
TO MY "FOREVER" FAMILY
Producing this dissertation involved the skills of many people
and, although I accept full responsibility for its contents, I do not
claim all the credit.
I am indebted to the United States Army for providing the
opportunity to achieve this long time goal. I gratefully acknowledge
the selfless contributions of members of my graduate committee. Dr.
Jerry F. Butler, chairman, provided continuous instructional guidance,
moral and material support throughout the course of this study. Dr.
David G. Young was a true mentor, providing many unanswered questions
and the principal direction for my research; his enthusiasm for
natural science in general and medical entomology in particular was
very contagious. Dr. Stephen G. Zam gave freely of his time and
expertise in parasitology and protozoology and offered valuable
suggestions. I genuinely appreciate such a rare individual as Dr. A.
G. B. Fairchild for his wisdom, historical perspective, and his
ability as an editor to make kid gloves from a sow's ear. LTC(P) John
Reinert served on the committee for the first year and, in concert
with the other committee members, helped outline a demanding schedule
Sincere thanks go to Ms. Diana Simon, Mrs. Debra Boyd, and other
members of the laboratory staff for the many administrative and
logistical details I was able to take for granted. I extend special
thanks to Ms. Edna Mitchell for long hours of tedious work devoted to
maintenance of sand fly and rodent colonies, and to conducting life
cycle experiments. Mr. Ray Hale and Ms. Gabrielle Hodson assisted
with the photography and graphics, and Mrs. Margo Duncan offered
valuable suggestions in designing tables and figures. Ms. Debra Akin
generously shared her time and technical expertise in electron
microscopy. Mrs. Adele Koehler helped type the final manuscript.
I am grateful for the friendship of fellow graduate students and
cohorts, Dr. Richard Endris, Dr. Peter. Perkins, Dr. Richard Johnson,
CPT Terry Klein, Mr. Eric Milstrey, Mr. Bruce Alexander, and Mr.
Charles "Ben" Beard. Their suggestions and encouragement were
appreciated and their humor was therapeutic.
Finally, I wish to express my love and deepest gratitude to my
wife, Joyce, for her unwaivering love, support and sacrifice, and for
periodically reminding me of my true priorities. I thank my wonderful
children, Natalie, Juliet, Nathan, Joshua, Aaron, and Andrew, for the
three year's worth of afternoons, evenings, and weekends that could
have been theirs.
TABLE OF CONTENTS
ACKNOWLEDGEMENTS.... ......................................... iii
LIST OF TABLES.............................................. vii
LIST OF FIGURES................ .......... ..... .... .......... ix
1 SAND FLIES AND LEISHMANIASIS............................ 1
Introduction .......................................... 1
Historical Review of the Incrimination of Major
Vectors of Leishmaniasis.............................. 2
Old World Leishmaniasis............................ 7
New World Leishmaniasis........ .................. 13
Leishmaniasis in the United States of America....... 19
Statement of Objectives................................ 24
2 SAND FLIES ASSOCIATED WITH HUMAN CUTANEOUS LEISHMANIASIS
IN TEXAS: OBSERVATIONS ON THEIR BIOLOGY WITH SPECIAL
REFERENCE TO Lutzomyia diabolica (Hall).................. 26
Introduction ........................................ ... 26
Human Case Histories............................... 27
Description of Study Sites.......................... 33
Materials and Methods................................... 43
Determination of Sand Fly Fauna..................... 43
Processing and Maintenance of Wild-Caught Sand
Flies and Recovery of Eggs........................ 47
Results................................................ .. 49
Species of Sand Flies Present....................... 49
Processing and Maintenance of Wild-Caught Sand
Flies and Recovery of Eggs........................ 56
Accessory Glands and Parity......................... 58
Natural Parasite Infections......................... 61
Discussion and Conclusions. ............................. 66
Determination of Sand Fly Fauna..................... 66
Potential Vectors ................................. 81
Processing and Maintenance of Wild-Caught Sand
Flies and Recovery of Eggs........................ 81
Accessory Glands and Parity........................ 83
Natural Parasite Infections......................... 86
3 LABORATORY COLONIZATION OF Lutzomyia diabolica
WITH NOTES ON ITS BIOLOGY IN THE LABORATORY
(DIPTERA: PSYCHODIDAE) ................................. 91
Introduction........................... .. .......... 91
Materials and Methods................................... 92
Immature Stages..................................... 93
Adults ................... .......... ...... ........ 97
Results .................................................. 100
General Observations .............................. 100
Immature Stages..................................... 103
Age-Specific Life Table........................... 134
Discussion and Conclusions.............................. 136
General............... ...... ...... ..... .............. 136
Immature Stages.................................... 137
Adults....................... .................... 151
Age-Specific Life Table........................... 157
4 EXPERIMENTAL TRANSMISSION OF Leishmania mexicana BY
BITES OF Lutzomyia diabolica (HALL) AND Lutzomyia
shannoni (DYAR) WITH NOTES ON THE EXTRINSIC
DEVELOPMENT OF THE PARASITE............................. 162
Introduction.................................... ... 162
Materials and Methods.................................... 163
Infection of Sand Flies and Parasite Development.... 163
Ultrastructure Studies............................. 166
Transmission Trials................................. 167
Results ................................................ 168
Infection of Sand Flies and Parasite Development.... 168
Ultrastructure Studies.............................. 188
Transmission Trials ................................. 190
Discussion and Conclusions ............................... 206
Infection of Sand Flies and Parasite Development.... 206
Ultrastructure Studies.............................. 215
Transmission of Leishmaniasis....................... 218
5 SUMMARY AND RECOMMENDATIONS.............................. 226
BIOGRAPHICAL SKETCH......................................... 243
LIST OF TABLES
1-1. Proven or suspected sand fly vectors of leishmaniasis
in the New World.................................... 3
2-1. Summary of Lutzomyia sand fly surveys conducted
in vicinities of human case sites of cutaneous
leishmaniasis in south central Texas................... 50
2-2. Number of male and female Lutzomyia diabolica in
selected collections from three sites using six
trapping methods............................. ........ 51
2-3. Fecundity summary of wild-caught Lutzomyia diabolica
females from south central Texas....................... 59
2-4. Summary of preoviposition intervals, postcapture and
postoviposition longevities of wild-caught Lutzomyia
diabolica females from south central Texas............. 60
2-5. Incidence of natural parasite infections in 341 female
Lutzomyia diabolica collected in south central
Texas ................................................ 64
3-1. Summary of fecundity, incubation time, and hatching
for generations 1 through 9 and 13 in a laboratory
colony of Lutzomyia diabolica.......................... 105
3-2. Comparison of life-cycle parameters of immature
Lutzomyia diabolica reared at 240C and 270C............ 107
3-3. Duration of larval and pupal stages in generations 1
through 9 and 13 of a laboratory colony of Lutzomyia
3-4. Comparison of immature development times for male and
female Lutzomyia diabolica reared individually to
adult stage at 24oC and 27C ........................... 114
3-5. Comparison of life-cycle parameters of sand flies
(Lutzomyia diabolica) reared on 5 larval diets......... 117
3-6. Summary of oviposition and longevity data for
generations 2 through 9 and 13 in laboratory colony
of Lutzomyia diabolica................................ 130
3-7. Age-specific life table for laboratory-reared
Lutzomyia diabolica based on observations of
several generations.................................. 135
4-1. Infection rates in laboratory-reared Lutzomyia
diabolica and Lu. shannoni sand flies fed on
hamsters with hTstiocytomas due to Leishmania
mexicana (strain WR-411).............................. 169
4-2. Distribution of Leishmania mexicana (strain WR-411)
in the alimentary tract of Lutzomyia diabolica
dissected post mortem after feeding on infected
hamsters......... .... ........................... ....... 172
4-3. Distribution of Leishmania mexicana (strain WR-411)
in the alimentary tract of Lutzomyia shannoni
dissected post mortem after feeding on infected
hamsters..................... .................... 173
4-4. Description and location of Leishmania mexicana
(strain WR-411) in the alimentary tract of laboratory-
reared Lutzomyia diabolica 12 hrs to 8+ days after
an infecting blood meal................................ 174
4-5. Distribution of Leishmania mexicana amazonensis
(strain untyped) in the alimentary tract of Lutzomyia
diabolica dissected post mortem after feeding
on histiocytomas of infected hamsters................ 185
4-6. Distribution of Leishmania braziliensis guyanensis
(strain MHOM/SR/80/CUMC 1) in the alimentary tract of
Lutzomyia diabolica dissected post mortem after
feeding on a histiocytoma on the tail of a white
4-7. Results of transmission trials of Leishmania
mexicana (strain WR-411) by bites of laboratory-
reared Lutzomyia diabolica........................... 196
4-8. Results of postfeeding dissections of laboratory-
reared Lutzomyia diabolica sand flies involved in
experimental transmissions of Leishmania mexicana
(strain WR-411) to Syrian hamsters.................... 200
4-9. Results of transmission trials of Leishmania
mexicana (strain WR-411) by bites of laboratory-
reared Lutzomyia shannoni............................. 201
4-10. Results of postfeeding dissections of laboratory-
reared Lutzomyia shannoni sand flies involved
in experimental transmissions of Leishmania mexicana
(strain WR-411) to Syrian hamsters.................... 203
LIST OF FIGURES
2-1. Distribution of autochthonous human cases of
cutaneous leishmaniasis in Texas and locations
of study sites......................................... 28
2-2. Cutaneous lesion due to Leishmania mexicana on
ear lobe of patient B............................................ 31
2-3. Cutaneous lesion due to Leishmania mexicana on
thigh of patient C..................................... 31
2-4. Cutaneous lesion due to Leishmania mexicana on
cheek of patient D .................................... 32
2-5. Habitat typical of that found at Garner State Park,
Concan, Uvalde County, Texas.......................... 38
2-6. Habitat at Garner State Park, Concan, Uvalde County,
Texas, showing public latrine resting station.......... 38
2-7. Habitat typical of that found at Seminole Canyon
State Park, Val Verde County, Texas.................... 39
2-8. Habitat at Fawcett Boy Scout Camp, Barksdale,
Edwards County, Texas, showing open latrine resting
station........................................ ....... 39
2-9. Farmland surrounding the home of patient D in the
rural community of D'Hanis, Medina County, Texas....... 40
2-10. Hunting dogs in a kennel behind the home of
patient D in D'Hanis, Medina County, Texas............. 40
2-11. Wood pile located behind the home of patient D in
D'Hanis, Medina County, Texas.......................... 41
2-12. Hog pen located about 80 m west of the home of
patient D in D'Hanis, Medina County, Texas............. 41
2-13. Romer ranch, near Devine, Medina County, Texas,
home of the paternal grandparents of patient B......... 42
2-14. Neighborhood of patient B in northwest San
Antonio, Bexar County, Texas........................... 42
2-15. Neighborhood of patient A in southeast San
Antonio, Bexar County, Texas.......................... 44
2-16. Shannon trap erected behind the home of patient D,
D'Hanis, Medina County, Texas.......................... 46
2-17. Lutzomyia diabolica female............................ 53
2-18. Dismantled woodrat (Neotoma) nest near the home of
patient D in D'Hanis, Medina County, Texas, where
specimens of Lutzomyia anthophora were found........... 57
2-19. Paired accessory glands of a gravid female Lutzomyia
diabolica showing dense granular material.............. 62
2-20. Condition of accessory glands in relation to number
of eggs deposited.................................... 63
2-21. Unidentified flagellates in the midgut of a
Lutzomyia diabolica female collected at Garner
State Park, Uvalde County, Texas, June 1982............ 67
2-22. Gamont of aseptate gregarine in the abdominal
cavity of a female Lutzomyia diabolica collected
at Garner State Park, Uvalde County, Texas, June 1982.. 67
2-23. Gametocyst of an aseptate gregarine attached to
the accessory gland of a female Lutzomyia
diabolica collected at Garner State Park,
Uvalde County, Texas, June 1982....................... 68
2-24. Gregarine oocysts spilling from the accessory glands
of a female Lutzomyia diabolica collected
at Garner State Park, Uvalde County, Texas,
June 1982............................................. 68
2-25. Microsporidian spores in the hemocoel of a female
Lutzomyia diabolica collected at Garner
State Park, Uvalde County, Texas, June 1982........... 69
2-26. Microsporidian spore germinated in vitro by
addition of 0.2 M KC1 (pH 9) to T ~secting medium....... 69
2-27. Scanning electronmicrograph of mite (Eustigmaeus
sp.) found on the abdomen of a female Lutzomyia
diabolica collected at Garner State Park,
Uvalde County, Texas, June 1982........................ 70
2-28. Scanning electronmicrograph of mite (Eustigmaeus
sp.) found on the abdomen of a female Lutzomyia
diabolica collected at Garner State Park,
Uvalde County, Texas, June 1982........................ 70
2-29. Known geographical distribution of the sand fly
Lutzomyia diabolica in Texas........................... 73
2-30. Known geographical distribution of the sand fly
Lutzomyia anthophora in Texas......................... 80
2-31. Known geographical distribution of the sand fly
Lutzomyia texana in Texas.............................. 80
3-1. Bloodfeeding female sand flies, Lutzomyia
3-2. Number of adult progeny per parent female of
Lutzomyia diabolica reared during the first nine
generations of a laboratory colony..................... 101
3-3. Total number of Lutzomyia diabolica adults per
generation in a laboratory colony...................... 102
3-4. Generation times for generations 1 through 9 and 13
in a laboratory colony of Lutzomyia diabolica.......... 104
3-5. Life cycle of the Phlebotomine sand fly
Lutzomyia diabolica (Hall)............................. 110
3-6. Survivorship curves for immature Lutzomyia
diabolica reared at 240C and 270C...................... 113
3-7. Graphic comparison of immature development times
for male and female Lutzomyia diabolica reared
individually to the adult stage at 24C and 270C....... 115
3-8. Graphic comparison of immature development times for
Lutzomyia diabolica reared on five diet regimens....... 118
3-9. Proportion of diapausing eggs in batches laid
outdoors by Lutzomyia diabolica females between
June and December at Gainesville, Florida............. 120
3-10. Duration of the egg stage in outdoor-reared
Lutzomyia diabolica according to month of
oviposition at Gainesville, Florida.................... 121
3-11. Winter diapause development and spring emergence
in 33 Lutzomyia diabolica egg batches
deposited outdoors at Gainesville, Florida, between
August and December in 1982 and in 1983................ 123
3-12. Adult emergence patterns of Lutzomyia diabolica
in a laboratory colony (1st and 13th colony
generations)................... .................... 128
3-13. Adult emergence patterns of Lutzomyia diabolica
reared individually at 24C and 27............... ... 127
3-14. Emergence patterns of 12th generation Lutzomyia
diabolica adults reared at 270C and 87% RH
on five larval diets.................................. 128
4-1. Histiocytoma on the ear of a hamster infected with
Leishmania mexicana................................... 165
4-2. Infecting sand flies in a three-sleeved isolation
chamber by feeding them on a leishmanial
histiocytoma on a hamster's ear........................ 165
4-3. Diagram of a sand fly alimentary tract................ 171
4-4. Development of Leishmania mexicana in
Lutzomyia diabolica.................................. 180
4-5. Long-slender promastigotes of Leishmania mexicana
swimming freely in the hind triangle of an
infected sand fly, Lutzomyia diabolica................ 181
4-6. Short-broad promastigotes ("haptomonads") of
Leishmania mexicana spewing into the dissecting
medium at the site of a rupture in the gut wall of
a sand fly, Lutzomyia diabolica....................... 181
4-7. Short-broad, dividing promastigotes ("haptomonads")
massed at the stomodeal valve in the gut of a
female sand fly, Lutzomyia diabolica, infected
with Leishmania mexicana............................... 182
4-8. Severed anterior aspect of the cardia in a female
sand fly, Lutzomyia diabolica, showing massive
numbers of Leishmania mexicana promastigotes
("haptomonads") attached to the stomodeal valve........ 182
4-9. Long-slender, nondividing, binucleated prornastigote
seen in a 5-day infection of Leishmania
mexicana in Lutzomyia diabolica........................ 183
4-10. Extremely long mononucleated form seen in a 5-day
infection of Leishmania mexicana in
Lutzomyia diabolica.................................. .... 183
4-11. Large peanut-shaped flagellate observed in infected
and uninfected laboratory-reared Lutzomyia
diabolica females............................ ... .... 187
4-12. Electronmicrograph of Leishmania mexicana
anastigotes in large vehicles within a macrophage
cell of a hamster infected by the bite of a
female Lutzomyia diabolica ............................ 189
4-13. Electronmicrograph of saggital and transverse
sections of Leishmania mexicana amastigotes in
macrophage cells of an infected hamster................ 192
4-14. Electronmicrograph of Leishmania mexicana
promastigotes in the abdominal midgut of an
infected Lutzomyia diabolica female.................... 193
4-15. Electronmicrograph of Leishmania mexicana
pronastigotes in the cardia (thoracic midgut) of
an infected Lutzomyia diabolica female................. 195
4-16. Lesion on the ear of a hamster approximately two
weeks after it was bitten by a female Lutzomyia
diabolica that was experimentally infected with
Leishmania mexicana................................. 199
Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
BIOLOGY AND COLONIZATION OF THE SAND FLY
Lutzomyia diabolica (HALL) (DIPTERA: PSYCHODIDAE)
WITH NOTES ON ITS POTENTIAL RELATIONSHIP TO HUMAN
CUTANEOUS LEISHMANIASIS IN TEXAS, USA
Phillip G. Lawyer
Chairman: Dr. Jerry F. Butler
Cochairman: Dr. David G. Young
Major Department: Entomology and Nematology
A survey for potential sand fly vectors of human cutaneous
leishmaniasis was conducted within an endemic focus of the disease in
south central Texas, USA. Five species of Lutzomyia, including one
new species, were collected and eight new county records established.
Lutzomyia diabolica (Hall), the only anthropophilic sand fly
encountered, was the most commonly collected, accounting for 99% of
the total catch. This species was taken in light trap collections
throughout the frost-free season at a case site in D'Hanis, Texas. No
natural leishmanial infections were observed in more than 600
dissections of wild-caught female Lu. diabolica. Several natural
infections of nonleishmanial parasites are reported for the first
The first productive laboratory colony of Lutzomyia diabolica was
established and detailed studies of the fly's biology were conducted
through 16 generations. A new larval diet, developed from a modified
horn fly [Haematobia irritans (Linn.)] diet, reduced the average
immature development time by 50%, to about 33 days. Quiescence and
diapause occurred in both the egg and larval stages, and lasted as
long as 270 days in the egg stage of outside-reared sand flies.
Infection rates of 88% and 95% in Lu. diabolica and Lu. shannoni,
respectively, were obtained by feeding the flies on leishmanial
histiocytomas of laboratory-infected hamsters. The development of
Leishmania mexicana (strain WR-411) in Lu. diabolica is described in
Ultrastructure studies of L. mexicana amastigotes and
promastigotes revealed that the subpellicular microtubule number
varies widely and is not a good criterion for distinguishing the
For the first time, transmission experiments demonstrated that Lu.
diabolica and Lu. shannoni are able to transmit L. mexicana to hamsters
by individual bites. Transmission occurred after a period of parasite
multiplication and development, during which time massive infections
were established initially in the midgut, then in the cardia and at
the stomodeal valve. These gave rise to short-slender, highly active
promasitgotes that spread throughout the alimentary tract and invaded
the pharynx and mouthparts.
Based on the findings of this study, Lutzomyia diabolica is the
suspected vector of human cutaneous leishmaniasis in Texas.
SAND FLIES AND LEISHMANIASIS
The common name "sand fly" has been used confusingly in the
literature for members of two families of Diptera, the
Ceratopogonidae and the Psychodidae. In this work "sand fly"
refers only to the members of Psychodidae belonging to the subfamily
Sand flies are important as vectors of several human pathogens
including phleboviruses (e.g., sand fly fever), bartonellosis
(Carrion's disease) and, most notably, leishmaniasis, a complex of
diseases caused by various species and subspecies of unicellular
hemoflagellates in the genus Leishmania Ross (Adler and Theodor,
1957). Leishmaniasis is widely distributed in most tropical and
subtropical countries, extending through Central and South America,
Central and Southeast Asia, India, China, the Mediterranean Basin, and
Africa (Lainson, 1982). Until recently the disease was believed to be
absent from North America north of Mexico, but the confirmation of
several autochthonous human cases in Texas since 1968 has dispelled
that belief (Simpson et al., 1968; Shaw et al., 1976; Gustafson et al.,
1984). In 1981 the World Health Organization (WHO) estimated that 400
thousand new cases of leishmaniasis occur annually throughout the
world, but this may be an underestimation. Leishmaniasis is probably
second in importance only to malaria among the protozoan diseases in
terms of human suffering and economics (Lainson, 1982). Selection of
leishmaniasis by WHO as one of six diseases of man warranting a
special program of study has led to increasing interest in sand flies,
the only known natural vectors of the disease (Killick-Kendrick, 1978;
WHO, 1981). It should be noted that workers in Oklahoma fed nymphal
ticks [Rhipicephalus sanguineus (Latrielle)] on dogs infected with L.
infantum and found that the ticks remained culture positive for the
parasite for at least one month postmolting (Fox, pers. comm., 1984).
Culture-positive adult ticks that subsequently fed on uninfected
puppies, transferred Leishmania to the puppies. Although this
demonstrates that ticks can experimentally transmit leishmaniasis, it
may not occur under natural circumstances.
Previous work on sand flies has been mostly taxonomic. More
recently, however, interest has turned to the biology of sand flies,
the most important practical aspect being their relationship with
leishmanial parasites. The role as vectors of most of the 53 species
or subspecies of sand flies thought to transmit leishmaniasis to man
requires confirmation (Killick-Kendrick, 1978). Of the 21 species of
New World sand flies reported or suspected as being natural hosts of
Leishmania spp. infecting man, only six have been definitely
incriminated as vectors (Table 1-1). Prior to this study no
anthropophilic species in the USA had been incriminated in the
transmission of leishmaniasis.
Historical Review of the Incrimination of
Major Vectors of Leishmaniasis
Cutaneous and visceral leishmaniases are apparently ancient
afflictions of man. As early as the first century AD, in Central
Proven or suspected sand fly vectors of leishmaniasis in
the New World, by country (Killick-Kendrick, 1978; WHO,
Leishmania Lutzomyia sand fly
Country parasites vectors
Table 1-1. Continued.
Leishmania Lutzomyia sand fly
Country parasites 1 vectors
L. d. c.
L. d. c.
L. d. c.
L. b. p.
L. b. p.
L. b. p.
L. b. p.
L. d. c.
L. b. g.
L. d. c.
L. m. a.
L. m. g.
Leishmania species: L. m. m. = L. mexicana mexicana, L. d. c. =
L. donovani chagasi, L. b. b. = L. braziliensis braziliensis, L. b. g. =
L. braziliensis guyanensis, L. m. = L. mexicana, L. b. p. = L. brazil-
Tensis panamensis, L. e. = L. peruviana, L. m. a. = L. mexicana
amazonensis, L. m. 1. = L. mexicana garnhami.
Asia, the cutaneous disease was referred to as "Balkh Sore" (after a
town in north Afghanistan, near the Russian border), and by early
travelers as "Aleppo Boil" in Syria and "Baghdad Boil" in Iraq
(Lainson, 1982). In the Americas, Peruvian and Ecuadorean pottery
from the era 400 to 900 AD depicts human faces with mutilations very
similar to those caused by cutaneous and mucocutaneous leishmaniasis.
Spanish historians at the time of the conquest described severely
mutilating sores on the faces of Peruvian Indians (Lainson, 1982).
Because the major signs and symptoms of visceral leishmaniasis
resemble those of several other tropical diseases, it is difficult to
trace clear-cut references to this disease in ancient writings in
either the Old or the New World. The disease first attracted public
attention in 1882 when Clark, of the Sanitary Commission of India,
gave an account of 100 cases of a severe form of malarial cachexia,
depopulating areas of the Garo Hills, Assam. Natives of the area
called the disease "kala-azar" (black fever) and it appears that it
was known to them as early as 1869 (Strong, 1945). Epidemics of what
must have been the same disease, under the name of "Burdwan fever,"
occurred in lower Bengal from 1854 to 1875, causing a quarter of a
million deaths (Strong, 1945). During this same period in the
Mediterranean region it was referred to as "infectious splenic anemia"
or "infantile splenic anemia" (Lainson, 1982).
Visceral leishmaniasis apparently was unrecognized as a distinct
disease in Latin America until the time of the first parasitologically
proven case, in Paraguay in 1913 (Migone, 1913). This fact, coupled
with clinical, epidemiological, and biochemical similarities between
Mediterranean and American visceral leishmaniasis, suggests that it was
introduced since the European discovery of the New World (Lainson,
Possibly the first written account implicating sand flies in the
transmission of human pathogens, including leishmaniasis, appeared in
1764 in a sort of almanac published in Lima, Peru,under the direction
of Cosme Bueno, distinguished physician, mathematician and geographer
(Herrer and Christensen, 1975). In El Conocimiento de los Tiempos he
discussed the folklore regarding the natural transmission of verruga
(bartonellosis) and corrosive facial ulcers (uta, a form of cutaneous
leishmaniasis), reporting that both diseases originate from the bite
of a small insect called "Uta" (sand fly). More than a century later,
Mitford reported on cutaneous leishmaniasis in the Middle East and
considered the possible participation of some insect in the transmis-
sion of Aleppo boil, although its exact role was not clearly indicated
(Lewis, 1978). In 1904 Rogers discovered that the causative agent of
Indian kala-azar (visceral leishmaniasis) developed into a leptomonad
flagellate in culture. He also noted that similar organisms (e.g.,
Leptomonas) had been found in insects (mosquitoes), tangentially sug-
gesting an insect vector of kala-azar (Rogers, 1904). The ensuing
search for vectors of kala-azar (Leishmania donovani) and oriental
sore (Leishmania tropica) included a wide range of suspects including
bed bugs, fleas, mosquitoes, house flies, sand flies, hippoboscids,
and even leeches. In 1905, the Sergents and Pressat, attracted by the
coincidental distribution of sand flies and leishmaniasis, indepen-
dently suggested that these insects were probable vectors of oriental
sore (Sergent et al., 1914; Kirk and Lewis, 1955). Wenyon (1912)
reviewed the advances made in the knowledge of leishmaniasis in
the Old and New Worlds and supported the suggestion that some sort
of insect was involved in its transmission. He and others
unsuccessfully attempted to demonstrate leishmaniasis transmission
using fleas (Ctenocephalides canis and Pulex irritans), mosquitoes
(Stegomyia fasciata = Aedes aegypti), bed-bugs (Cimex sp.), and sand
flies (Phlebotomus sp.). On another occasion, Wenyon (1911)
dissected a number of wild-caught sand flies from Aleppo and observed
"Herpetomonas" flagellates in about 6% of the specimens; he
acknowledged the possibility that what-appeared to be harmless
parasites of sand flies might, in fact, be developmental forms of
L. tropica. Wenyon's discovery marked the beginning of intensified
efforts by numerous researchers to study all aspects of the parasitic
relationship between Leishmania and the sand fly host. For the next
30 years investigations progressed mainly on two fronts, in North
Africa and Palestine with oriental sore, and in India with kala-azar.
Old World Leishmaniasis
Leishmania tropica cutaneouss leishmaniasis, "oriental sore").
In a note on the etiology of oriental sore in Mesopotamia, Patton
(1919) believed that P. papatasi Scopoli and probably P. minutus
Rondani were carriers of the parasite. Acton (1919) showed that the
distribution on the body of oriental sores corresponded to the
distribution of bites by Phlebotomus. In 1921, Sergent et al.,
working in Algeria, first described the transmission of oriental sore
to a human. They divided 559 sand flies into 23 batches, crushed them
in saline and inoculated the resulting suspensions into the arms of 23
volunteers. The flies had been collected in Biskra, an endemic center
of the disease 600 km south of Algiers where the disease does not
occur. Only one of the inoculations produced positive results (Adler
and Theodor, 1925).
Adler and Theodor (1925), studying Phlebotomus in Jericho
(Jordan) found heavy "Herpetomonas" infections in four female sand
flies out of nearly 400 dissected. Material from one of these
infected flies was inoculated into the forearm of a volunteer and a
small papule containing Leishman-Donovan bodies (amastigotes) was
subsequently produced. Following this successful transmission of
oriental sore, they conducted feeding experiments to determine if
"Herpetomonas tropica" was capable of developing in P. papatasi after
a feed on an oriental sore, and whether or not the parasite remained
infective to man after passing through the sand fly. Dissections were
made two to seven days after the infective feed; 16 sand flies were
found to be infected. Material from these artificially infected flies
was inoculated into volunteers, none of whom subsequently showed signs
of infection. Although, Adler and Theodor felt that their experiments
had provided sufficient proof of the role of sand flies as vectors of
oriental sore, this view was criticized by Wenyon (Lainson, 1982), who
emphasized the need to transmit oriental sore by natural bite of the
sand fly. Adler and Theodor replied by ingeniously demonstrating the
capability of the sand fly to transmit Leishmania by bite, although it
was not from man to man (Adler, 1928). Phlebotomus papatasi were fed
through a rabbit-skin membrane on a culture of L. tropica, and eight
days later were allowed to feed through another membrane on sterile,
inactivated rabbit serum. Some of this serum was sown into blood-agar
medium used for culturing Leishmania. Flagellates were
present six days later. The same authors may have actually
transmitted Leishmania to man by bite of the sand fly one year later.
They fed experimentally infected P. sergenti Parrot on a number of
volunteers and obtained a positive lesion on the arm of one man. It
should be noted that the incubation period was so long that the
individual had visited endemic areas of oriental sore before
appearance of the lesion and therefore could possibly have acquired
the infection there (Adler and Theodor, 1929). Finally, 20 years
after Sergent and colleagues artificially transmitted oriental sore by
scarification, Adler and Ber (1941) proved without question that
artificially infected P. papatasi can transmit oriental sore to a
human by bite. The key to this success appears to have been in
feeding the sand flies on flagellates suspended in three parts 2.7
percent saline and one part defibrinated blood.
Leishmania donovani (visceral leishmaniasis, "kala-azar"). In
1915 Mackie, convinced that a relationship existed between kala-azar
and some biting insect, attempted to make a hut to hut insect census
in kala-azar-infected villages. His team collected body lice, head
lice, bed bugs, mosquitoes, sand flies and even leeches in the bedding
or on the persons of patients who were proven to have active kala-
azar; these specimens were carefully examined for Leishmania
parasites. All specimens were negative except the sand flies which
contained "Herpetomonas parasites" (probably Leishmania), "bodo-like
parasites," and "sporozoan-like parasites." Not realizing how close
he must have been to linking sand flies with kala-azar, Mackie stated:
"This long series of negative results rather tends to check enthusiasm
for the insect-borne hypothesis of kala-azar. The only insect
which has given any return for work spent on it is the sand fly and I
am of the opinion that the relation of this insect to disease would
repay further investigation" (1915, p. 949).
Perhaps stimulated by Mackie's suggestion and following the
example of Sergent et al. (1921) in their work on oriental sore
in Algeria, Sinton (1922) studied the distribution of Indian sand
flies and found that the distribution of kala-azar showed striking
congruence with that of P. argentipes Annandale and Brunetti; he
considered this sand fly to be the most likely vector. In 1924, a
"Kala-azar Commission" was set up under the directorship of
Christophers, and its members were soon successful in rearing this
most likely suspect in the laboratory. Once a laboratory colony was
established, Knowles et al. (1924) fed specimens of these sand flies
on patients with kala-azar and found that a heavy flagellate infection
developed in P. argentipes; they noted that the same degree of
infection did not occur in other Phlebotomus species. The following
year Shortt et al. (1926), working with L. donovani, demonstrated that
in P. argentipes the infection passes forward in the fly to the buccal
cavity and proboscis. For the next five years, Shortt and co-workers
of the Kala-azar Commission conducted three series of transmission
experiments using human volunteers and Chinese hamsters as hosts.
They fed an astounding 79,939 artificially infected P. argentipes on
either human volunteers or hamsters but reported negative results as
The failure of any of the subjects of experiment to show
infection with kala-azar is very difficult to explain if
the theory of Phlebotomus transmission is to be maintained. We
can only suppose that some essential factor in the process of
infection has been omitted in our experiments which is present
under natural conditions, or that the vast amount of labour
expended by us during a period of 5 years has been expended
on an insect which is not an essential link in the chain of
infection. (Shortt et al., 1930, p. 929)
It appears that this final report was written before all the results
were in, for on February 19, 1931, the following telegram from New
Delhi was received by Nature magazine: "Lieut-Col. Shortt reports
successful transmission of Leishmania donovani to Chinese hamsters by
bites of artificially infected Phlebotomus argentipes. Hamster bitten
repeatedly during twelve months; generalized infection found seventeen
months after experiment began" (Shortt, 1931, p. 308). Their
persistence had paid off.
The kala-azar commission continued its efforts to transmit kala-
azar to man by bite of P. argentipes but met with uniformly negative
results (Swaminath et al., 1942). It was only after Smith et al.
(1940) devised the technique of keeping the flies alive after
oviposition by feeding on the juice of boiled raisins that Swaminath
et al. (1942), using this technique, were able to forge the final link
of evidence incriminating P. argentipes as the insect vector of kala-
azar in India. They suggested that feeding the flies on fruit juices
acted either by increasing the virulence of the parasites or
increasing the parasitemia, thus enabling them to reach the anterior
part of the midgut (cardia) more rapidly.
Leishmania donovani infantum (infantile visceral leishmaniasis).
The coincidental distributions of visceral leishmaniasis and the sand
fly P. chinensis Newstead in China north of the Yangtze River, pointed
to this insect as the vector of the disease. Young and Hertig (1926),
in North China, dissected hundreds of field-caught P. chinensis, P.
sergenti, and P. perturbans de Meijere and examined them for the presence
of flagellates; all specimens were negative. They also attempted to
incriminate these three species by experimentally infecting laboratory-
bred sand flies, feeding them on kala-azar patients or infected
hamsters, and allowing them to refeed on uninfected hamsters. The
rates of infection were 85.3%, <2%, and 0% for P. chinensis,
P. sergenti, and P. perturbans, respectively. A small percentage of
these infected flies took a second blood meal, but transmission was
unsuccessful. Similar studies were conducted by Sun et al. (1936) and
Sun and Wu (1937) in which 7 of 21 P. chinensis, collected in houses
with cases of visceral leishmaniasis, were infected with
promasitgotes. Successful transmission to hamsters by the bite of
P. chinensis was finally reported by Feng and Chung (1941).
In the Cevennes, in southern France, Rioux and colleagues
accumulated overwhelming epidemiological evidence regarding the
suspected role of P. ariasi Tonnoir as the vector of infantile
visceral leishmaniasis. Not until 1979, however, did they confirm
their suspicions by transmitting the disease to a dog by the bite of
an experimentally-infected sand fly (Rioux, et al., 1979).
Leishmania major cutaneouss leishmaniasis, wet sore). It is
unclear when L. major was first transmitted experimentally by bite of
a sand fly because early workers recognized only one leishmanial
species that was divided into two subspecies, "minor" and "major"
causing "dry" and "wet" oriental sore, respectively. Perfil'ev (1968)
claimed that the first successful transmission of cutaneous
leishmaniasis by bite of a sand fly was accomplished in 1941 by
Kryukova in experiments with gerbils. Laboratory-bred P. papatasi
were infected with cutaneous leishmaniasis by feeding on a histocytoma
on a gerbil's ear and were subsequently allowed to take a second blood
meal from an uninfected gerbil. Lesions developed on the second
gerbil 15 days post infective feed. In 1927 Koshevnikova and
coworkers showed that the principal reservoir host of L. tropica in
Central Asia is man and that the host of L. major is the gerbil
(Rhombomys opimus) (Perfil'ev, 1968). Therefore, it seems safe to
assume that Kryukova's results represent the first experimental
transmission of L. major by the bite of a sand fly, because the
original pathogen was obtained from gerbils caught in Turkmenia.
New World Leishmaniasis
In the New World, the task of vector incrimination has been
complicated by three factors: the extremely wide variety of sand fly
species (about 327 Lutzomyia compared to about 101 Phlebotomus in the
Old World), the failure to appreciate the multiplicity of leishmanial
parasites on the part of some workers, and the difficulty of working
in a dense tropical rain forest (Lainson, 1982).
Leishmania donovani chagasi (American visceral leishmaniasis).
The peridomestic nature of American visceral leishmaniasis due to
L. donovani chagasi facilitated the incrimination of Lu.1 longipalpis
(Lutz and Neiva) as the main vector of this disease. Its geographic
distribution coincides with visceral leishmaniasis throughout Latin
In 1936, Evandro Chagas found Lu. longipalpis in the house of the
first case of visceral leishmaniasis to be studied in South America,
in Sergipe, Brazil (Lainson, 1982). This prompted Chagas and others
1The abbreviation Lu. for Lutzomyia is used to avoid confusion with L.
to feed Lu. longipalpis on infected dogs. Promastigotes developed in
the guts of these sand flies and were infective upon subsequent
inoculation into uninfected hamsters (Lainson, 1982). With the
accidental death of Evandro Chagas in 1940, interest in visceral
leishmaniasis also seemed to die, not to be resurrected until 1954
when Deane and Deane, investigating serious outbreaks of the disease
in the state of Ceara, Brazil, found wild-caught Lu. longipalpis
heavily infected with promastigotes believed to be L. donovani. They
also noted highly active flagellates in the biting mouthparts of other
Lu. longipalpis they had fed on a naturally-infected fox eight days
previously (Lainson, 1982). Although the promastigotes seen by the
Deanes were not proven to be L. d. chagasi, the epidemiological
evidence was so strong that there remained little doubt as to the
importance of this sand fly (Lainson, 1982).
Further evidence to incriminate this peridomestic vector came in
1977 when Lainson and colleagues achieved five separate transmissions
of L. d. chagasi in hamsters, by bite of laboratory-bred insects which
had previously ingested amastigotes in artificially infected rabbit
blood (Lainson et al., 1977).
Leishmania mexicana (American cutaneous leishmaniasis, bay sore,
chiclero's ulcer). Experimental infection of sylvatic sand flies with
parasites causing American cutaneous leishmaniasis has been
accomplished in a wide range of species, but experimental transmission
has been difficult to achieve because of the wide choice of potential
vectors. Strangways-Dixon and Lainson (1962) infected wild-caught
females of nine species of sand flies in Belize by feeding them on
hamsters exhibiting skin lesions due to L. m. mexicana. The flies then
were fed on human volunteers, one of whom developed a small lesion 17
days after one insect made a 30 second bloodless probe. The vector
was initially identified as "P. paraensis" Costa Lima, but was later
reexamined by Williams (1983) and determined to be most similar to
Lu. panamensis (Shannon) rather than paraensis. To demonstrate a natural
infection of leishmaniasis in sand flies and to strengthen their case
for incriminating the vector (Lu. panamensis), Strangways-Dixon and
Lainson (1962), collected approximately 270 wild female sand flies,
triturated them in sterile Locke's solution and inoculated half of
the suspension into the back of an uninfected hamster. Two months
later a small dermal swelling was noted at the site of the
inoculation. Stained smears from the lesion revealed leishmaniae
(Strangways-Dixon and Lainson, 1962). This experiment showed that
some wild-caught sand flies were naturally infected, but since all
specimens collected were pooled, it was not determined which species
was (were) infected. Subsequent dissections of 334 newly caught, man-
biting Lutzomyia revealed epimastigote-like infections in two specimens,
one each of Lu. ovallesi (Ortiz) and Lu. cruciata (Coq.). In view of
the scanty infections, the authors felt that these more likely
represented Leptomonas or Herpetomonas rather than Leishmania
In Brazil, wild-caught Lu. longipalpis and Lu. renei (Martins,
Falcao, and da Silva), allowed to feed on a strain of Leishmania
isolated from a human patient in Belize, also successfully transmitted
the parasite (Coelho and Falcao, 1962).
Williams (1966a) collected wild sand flies in Belize and fed them
on a leishmanial lesion on a hamster. Two to three days later, the
flies fed on the forearm of a human volunteer. Two lesions were
produced about four weeks later, and in both cases the vector fly was
In the same country, Disney (1966) designed a trap to catch sand
flies attracted to rodents and found that Lu. olmeca olmeca (Vargas
and Najera) was highly attracted to them. Later, using this
technique, he found Lu. o. olmeca naturally infected with L. m.
mexicana (Disney, 1968). In neighboring Yucatan Peninsula, Biagi et
al. (1965) confirmed the importance of Lu. o. olmeca as a vector of
leishmaniasis, and transmitted the parasite to a volunteer by bite of
a naturally infected fly (Lainson, 1982).
Lainson and Shaw (1968) demonstrated that the vector of
L. mexicana amazonensis in the Amazon forests of Brazil is
Lu. flaviscutellata (Mang.), a species highly attracted to the rodent
Proechimys guyanensis, the principal reservoir host. Of 7,322 flies
dissected between 1968 and 1973, 45 or 0.6% were infected with
promastigotes. Parasites from 18 of these infected flies were
inoculated into hamsters and 15 of the inoculations produced typical
L. mexicana amazonensis infections (Ward et al., 1973). These workers
also transmitted the disease from hamster to hamster on four occasions
using laboratory-bred Lu. flaviscutellata (Ward et al., 1977). From
these findings it was concluded that Lu. flaviscutellata is the
principal, and probably only, vector of L. mexicana amazonensis in the
Amazon region (Lainson and Shaw, 1979).
Leishmania brazilienensis panamensis (Panamanian cutaneous
leishmaniasis). Natural flagellate infections have been recorded in
numerous neotropical sand fly species, but it was not until extensive
studies were carried out at the Gorgas Memorial Laboratory in Panama
that they were proven to be Leishmania (Hertig and McConnell, 1963;
Johnson et al., 1963). After many years of patient work, these
researchers found natural promastigote infections in more than 400
man-biting Panamanian sand flies and incriminated three species,
Lu. trapidoi (Fairchild and Hertig), Lu. ylephiletor (Fairchild and
Hertig) and Lu. gomezi (Nitz.) as vectors of Panamanian cutaneous
leishmaniasis, L. braziliensis panamensis (Johnson et al., 1963).
Christensen et al. (1969) also found Lu. panamensis infected with
promastigotes, but there is some question as to the identity of those
promastigotes. Lainson (1982) believed that some of these infections
were due either to Endotrypanum (a blood parasite of sloths which also
develops as promastigotes in sand flies) or to nonhuman Leishmania,
such as Lu. hertigi of porcupines.
McConnell (1963) cultured flagellates from wild-caught sand flies
and determined, without question, that Lu. trapidoi harbored
promastigotes of L. b. panamensis. Furthermore, the observation that
Lu. trapidoi is largely arboreal led to the incrimination of sloths as
the major reservoir of L. braziliensis panamensis (Lainson, 1982).
Leishmania braziliensis guyanensis ("pian-bios"). In Surinam,
Wijers and Linger (1966) caught large numbers of anthropophilic sand
flies off human bait in areas where "pian-bois," due to L. b.
guyanensis, is endemic. The most common species recorded was
Lu. squamiventris (Lutz and Neiva), but all dissections of this species
proved negative for leishmaniae. Numerous promastigote infections
were found in dissections of another species, "Lu. anduzei," found
resting on tree trunks. Their "anduzei" probably represented
Lu. umbratilis (Ward and Fraiha) since it was found resting on tree trunks
(Ward and Fraiha, 1977). However, the role of this sand fly as a
vector remained speculative, since attempts to infect a hamster with
the parasite failed.
Lainson et al. (1976), observed a 7% infection rate in "Lu.
anduzei" (now Lu. umbratilis Ward and Fraiha) taken mostly from large
tree trunks during their studies of the epidemiology of "pian-bois" in
the Monte Dourado region, Para State, Brazil. Intradermal inoculations
of the flagellates into hamsters produced infections in all cases.
The parasite was shown to be identical biologically and biochemically
with that causing the disease in man. Lu. umbratilis was subsequently
incriminated as the vector of L. b. guyanensis in the neighboring
state of Amazonas, Brazil, by Arias and Freitas in 1977. Lainson et
al. (1976) reported heavy promastigote infections in seven specimens
of Lu. whitmani (Antunes and Coutinho), another tree trunk-inhabiting
sand fly. Unlike Lu. umbratilis, this sand fly was not particularly
anthropophilic, however, and they suggested that its importance is
probably limited to secondary transmission among wild animal
Leishmania braziliensis braziliensis (mucocutaneous
leishmaniasis, "espundia"). Incrimination of sand fly vectors of L.
b. braziliensis has been more difficult than with the other American
forms because of the poor growth of the parasite in laboratory animals
and in culture media, and because of difficulties in establishing
productive colonies of suspected vectors for transmission experiments.
Forattini et al. (1972), however, successfully infected hamsters with
promastigotes (believed to be L. b. braziliensis) found in two
naturally infected sand flies, Lu. intermedia (Lutz and Neiva) and Lu.
pessoai (Coutinho and Barretto) in Brazil. The former species is found
in low, secondary forests and is known also to invade houses, while
the latter is essentially sylvatic, but has been collected in houses
up to 300 m from the forest edge. With this in mind, it was suggested
that cutaneous leishmaniasis in southern Brazil may have a
peridomestic transmission, the original source of the infection being
in nearby wooded areas (Lainson, 1982).
In Serra dos Carajas, Pard State,-North Brazil, where both
cutaneous and mucocutaneous leishmaniasis are serious public health
problems, Lainson et al. (1973) concluded that Lu. welcome
(Fraiha, Shaw, and Lainson) was a major vector to man. Attempts to
infect hamsters by inoculation with parasites isolated from naturally
infected Lu. wellcomei were largely unsuccessful, but it was shown
that the parasite was the same as that infecting man in the same area.
Lu. wellcomei was considered to be of particular importance because of
the avidity with which it attacks man both during the night and the
Leishmaniasis in the United States of America
Few studies have been conducted on the epidemiology of
leishmaniasis in the USA. Prior to 1976 leishmaniasis was not
generally thought to occur autochthonously in the USA, and
potential sand fly vectors were reported so rarely as to be considered
of little medical consequence.
McEwen (1914) reported the first case of leishmaniasis in the
USA, referring to it as "oriental sore." There is little doubt that
he was dealing with American cutaneous leishmaniasis since the patient
acquired the lesion while traveling in South America (Stewart and
Pilcher, 1945). At that early date, differentiation between oriental
sore and American cutaneous leishmaniasis had not been made, nor had
any vectors been incriminated.
In 1919, Parman drew attention to a "Phlebotomus" species in
Uvalde, Texas,and noted that it attacked man. He did not attempt to
implicate it in disease transmission. This fly, described and named
Phlebotomus (Brumptomyia) diabolicus by Hall in 1936, represented the
first confirmed anthropophilic species of the genus Phlebotomus (now
Lutzomyia) from the USA. Lindquist (1936) stated that this species
was not a major pest in Uvalde, but that it frequently caused some
annoyance to people in southwestern Texas. He described a
peridomestic fly that often entered dwellings and other buildings to
feed on man and domestic animals. At that time, species of
Phlebotomus had been incriminated in the transmission of Phlebotomus
fever but were only suspected to play a role in the transmission of
leishmaniasis. Consequently, the fly was considered to be of little
importance other than being a nuisance.
Prior to 1943, approximately 30 human cases of cutaneous
leishmaniasis (oriental sore) had been reported from the USA and
Canada (Dwork, 1942), none of which (with one possible exception), were
autochthonous. The possible exception was a case reported by Gelber
(1942) in a 53-year-old woman from southern California who had a
typical leishmanial lesion on her left cheek. This case was thought
to be autochthonous because the woman had not left the country for 13
years. But since she had made an earlier tour of Europe and the
Mediterranean, the possibility cannot be ruled out that she might have
contracted the disease outside the USA.
Only three cases of mucocutaneous leishmaniasis had been reported
in the USA prior to 1943, one of which was presumed to be
autochthonous (Stewart and Pilcher, 1943). Benedek (1940) reported a
case of mucocutaneous leishmaniasis in a man from Chicago, which he
considered the first autochthonous case of leishmaniasis in the USA.
However, there is some question as to the extent of the patient's
travel, and thus there is room for doubt.
Stewart and Pilcher (1943) reported on a case of cutaneous
leishmaniasis in a 6-year-old Mexican-American boy who lived on a
ranch near Alice, Texas and had never traveled more than 60 miles from
his home. The authors, obviously aware of the recent incrimination of
sand flies as the vectors of both visceral and cutaneous leishmaniasis
in the Old World, mentioned that three'species of "Phlebotomus"
(Lutzomyia) had been identified in the USA, all of which were found
within or near Texas. They astutely concluded that the apparent
rarity of the disease in the United States was probably not real and
that the return of military and civilian personnel from endemic
centers was further reason for keeping American leishmaniasis in mind.
This was probably the first truly autochthonous case of leishmaniasis
reported in this country, although Wenyon disputed the identity of the
parasite stating: "The microphotograph illustrating the paper is a
good one, but though suggestive of Leishmania, it is not absolutely
convincing" (1945, p. 712).
Addis (1945a) described a new species of sand fly from Texas,
"Phlebotomus anthophorus" (Lu. anthophora), from specimens collected
at Uvalde, Texas,in the same locality where Lu. diabolica was found.
The female flies were collected in the mornings while feeding on
domestic rabbits. Attempts to feed this fly on man were unsuccessful.
Packchanian (1946) reviewed distributions and habits of the six
species of sand flies then known to occur in the USA. He drew
attention to the fact that sand flies were known vectors of
leishmaniasis in the Old World and were known to have been infected
naturally or experimentally with leishmaniasis in numerous localities
in Latin America. He stated that in all probability the species found
in the USA represented potential vectors of leishmaniasis and that
this important problem remained to be investigated under rigid
In 1968, Simpson et al. reported the first well documented, and
undisputed, autochthonous case of leishmaniasis in the USA. The
patient was a 64-year-old Mexican-American woman from San Benito,
Texas, who had a 57 year history of chronically active, disseminated
anergic cutaneous leishmaniasis.
Shaw et al. (1976) described two more autochthonous human cases
of cutaneous leishmaniasis in Texas. The first occurred in 1972 in a
74-year-old woman from Dilworth, Gonzales County,and the second in a
56-year-old man from Kenedy, Karnes County. The first patient owned a
ranch a few miles from her home and visited it daily to feed the
cattle and some stray dogs. She related that she frequently saw
swarms of gnats and small black flies, but did not remember any
specific insect bites, especially in relation to her skin lesions.
Aside from short visits to northern Mexico she had had no foreign
travel (Shaw et al., 1976). The second patient lived in a primitive
shack in rather unsanitary conditions. He reported no travel outside
the USA except for two half-day visits to Nuevo Laredo, Mexico. The
authors conducted epidemiologic studies in the neighborhoods and
communities surrounding the two case sites but collected no sand
flies. They concluded that conditions may be proper in south central
Texas for arthropod-borne transmission of cutaneous leishmaniasis, and
that the suspected endemicity of the disease should be confirmed by
Interest in the epidemiology of leishmaniasis in Texas mounted
slowly until, in 1980, cutaneous leishmaniasis (L. mexicana mexicana)
was diagnosed in a 11-year-old boy from Uvalde, Texas, the same
locality in which the first confirmed anthropophilic species of sand
fly in the USA (Lu. diabolica) was found (Gustafson et al., 1984).
The boy was presumed to have contracted the disease near his home or
while on camping trips in south central Texas, since his travel had
In that same year, Anderson et al. (1980) reported endemic canine
leishmaniasis in dogs near Oklahoma City, Oklahoma. The parasite most
closely resembles L. donovani infantum (Kocan et al., 1984). These
instances further emphasize the need to study the epidemiology of
leishmaniasis in the USA.
In 1981, Perkins (1982) demonstrated for the first time that an
anthropophilic USA sand fly, Lu. shannoni, could be experimentally
infected with an indigenous strain of L. mexicana (strain WR-411,
Uvalde, Texas) by feeding them on histocytomas on infected hamsters.
Although transmission was not accomplished at that time, the ground
work was laid for further studies. Later that same year Endris et al.
(Endris, pers. comm., 1984) demonstrated the ability of the non-
anthropophilic, rodent-feeding sand fly, Lu. anthophora, to transmit
L. mexicana (strain WR-411, Uvalde, Texas) from infected to uninfected
hamsters by bite. This was the first report of a native USA species
of sand fly transmitting leishmaniasis by bite. The authors suggested
that L. mexicana could be maintained in a wild rodent population by
Lu. anthophora from which it could then be transmitted to man by other
sympatric anthropophilic sand flies such as Lu. diabolica.
Finally, in 1983, Gustafson reported three confirmed and one
suspected case of cutaneous leishmaniasis from south central Texas
(Gustafson et al., 1984).
The vectors of this disease in Texas are unknown and until the
present study, no attempts had been made to incriminate any
anthropophilic sand fly from areas of Leishmania endemicity in the
Statement of Objectives
This study was undertaken to investigate the life history and
biology of the sand fly Lutzomyia diabolica (Hall) and its possible
role in the transmission of human cutaneous leishmaniasis in Texas.
For comparison, the vector capacity of Lu. shannoni (Dyar), another
anthropophilic sand fly, was investigated in conjunction with that of
Lu. diabolica. The specific objectives of the study were to
1. conduct a field survey of potential vector sand flies in
vicinities of recent human case sites of leishmaniasis in Texas;
2. study the field biology of Lu. diabolica and collect wild
stock for a laboratory colony;
3. establish a productive colony of Lu. diabolica in the
4. study the biology and life history of Lu. diabolica under
5. investigate, under controlled laboratory conditions, the
vector capacity of Lu. diabolica, as compared with Lu. shannoni, for
6. examine by means of the electron microscope the morphology of
L. mexicana promastigotes found in the.sand fly vector.
SAND FLIES ASSOCIATED WITH HUMAN CUTANEOUS LEISHMANIASIS
IN TEXAS: OBSERVATIONS ON THEIR BIOLOGY WITH SPECIAL
REFERENCE TO Lutzomyia diabolica (HALL)
Recent reports of human cutaneous leishmaniasis acquired in south
central Texas strongly suggest that the disease is endemic there (Shaw
et al., 1976; Gustafson et al., 1984). Because sand flies are the
only known natural vectors of leishmaniasis, a knowledge of their
field biology and host associations is of paramount importance in
epidemiologic studies of this disease. Six species of sand flies are
known to occur in Texas; these are Lutzomyia anthophora (Addis), Lu.
californica (Fairchild and Hertig), Lu. diabolica (Hall), Lu. oppidana
(Dampf), Lu. texana (Dampf), and Lu. vexator (Coq.) (Young and
Perkins, 1984). Only one, Lu. diabolica, is known to be
anthropophilic. This field study is the first to investigate sand
flies in Texas associated with specific human case sites of cutaneous
The investigation was initiated in 1982 with a survey trip in
late spring and early summer (4-28 June) to south central Texas. The
principal study site was Garner State Park near Concan in Uvalde
County. Secondary sites included Rio Frio in Real County, Seminole
Canyon State Park in Val Verde County, and Fawcett Boy Scout Camp near
Barksdale in Edwards County. The field work was conducted in
cooperation with Drs. D. G. Young, G. B. Fairchild, and R. G. Endris,
all from the University of Florida, Gainesville, FL. A second survey
trip was taken in early fall of 1983 (19-30 September), the principal
field study site being in and around the rural community of D'Hanis in
Medina County. Secondary sites included Garner State Park, the Romer
Ranch south of Devine in Medina County, and two sites within the city
limits of San Antonio in Bexar County. During the second trip the
able assistance of Mr. T. Long, zoonoti.c technician, Region 9, Texas
State Health Department, was greatly appreciated. The objectives
1. conduct a survey of potential vector sand flies in the
vicinities of recent human case sites of leishmaniasis in Texas (A
case site, as used herein, is defined as the home environs of a
confirmed leishmaniasis patient, or localities where that individual
camped or otherwise visited within three months prior to the onset of
disease symptoms.); and
2. study the field biology of Lu. diabolica and collect wild
stock for a laboratory colony.
Human Case Histories
Eight autochthonous human cases of cutaneous leishmaniasis have
been reported in south central Texas, four of which have occurred since
1980 (Gustafson et al., 1984) (Fig. 2-1). These latter four were the
only cases investigated during this study and are described below. Por-
tions of the following unpublished case histories were graciously provided
by Dr. T. Gustafson, the investigating Texas State epidemiologist. The
S= autochthonous human case of
= case site and study site
** study site
El cases reported since 1980
Distribution of autochthonous human cases of cutaneous
leishmaniasis in Texas and locations of study sites.
remaining information was obtained through personal interviews with
patients or their family members. A complete discussion of the case
histories and results of serologic tests are provided by Gustafson et
Patient A. The patient was an 11-year-old white male from Uvalde,
Texas, who noticed an ulcerating lesion on his left cheek beginning
February, 1980. In May, 1980, a biopsy was performed which showed
amastigotes in dermal macrophages. Leishmania mexicana was cultured
from biopsy material that was sent to Walter Reed Army Institute of
Research. Patient A had never traveled outside the USA except for
occasional one-day trips to the border city of Ciudad Acuna, Mexico.
In the month prior to onset, he had participated in a camping trip at
Fawcett Boy Scout Camp in Edwards County. The family had no pets.
Patient B. The patient was a 56-year-old white female from a
suburban neighborhood in southeast San Antonio, Texas, who first
noticed a small lesion on her left ear in November 1982 (Fig. 2-2).
She distinctly remembered waking one morning with an itching ear and
noticed a drop of blood on her ear lobe and on her pillow. She
suspected that she had been bitten by an insect. Her bed was located
next to an open, screened window. A biopsy performed in February,
1983, showed amastigotes in dermal macrophages by light and electron
microscopy. Patient B had traveled out of the USA as a military
dependent more than ten years previously and had visited Chihuahua in
northern Mexico in June, 1982. She had one pet dog.
Patient C. The patient was a 5-year-old white male from a
suburban neighborhood in northeast San Antonio, Texas, who first
noticed an enlarging papule on his left thigh beginning in November
1982 (Fig. 2-3). When the lesion persisted, an excisional biopsy was
performed which showed amastigotes in large vesicles within
macrophages by both light and electron microscopy. The patient had
never traveled outside the USA. He frequently spent the week ends at
the ranch of his paternal grandparents in Devine, Texas. The family
had one pet dog, and the boy's grandparents had three hunting dogs.
Patient D. The patient was a 10-year-old white male from the
rural farming community of D'Hanis, Texas, who developed a lesion near
his right eye brow and another on his right cheek in December, 1982
(Fig. 2-4). When the lesions persisted, an excisional biopsy was
performed which showed amastigotes in large vesicles within
macrophages by light microscopy. Presence of kinetoplasts was
confirmed by electron microscopy. Promastigotes were recovered from
an aspirate of the eyebrow lesion, but these died before further
identification could be performed. The patient had never traveled
outside of Texas and had lived all of his life in Medina County with
only occasional trips to Uvalde, San Antonio, and Seguin, Texas. He
regularly accompanied his father on coyote hunting trips near his
home. The family had ten hunting dogs. The patient's mother recalled
seeing tiny hopping flies in the house, around the light fixture, and
members of the family reported being bitten by tiny gnats. When shown
a live female Lu. diabolica the mother said that it was the same as
those she had seen in the house.
A suspected case, with history very similar to the others, was
reported from Hondo, Medina County, Texas, within 12 miles of the home
of Patient D. The patient was a 4-year-old white male who first
noticed a lesion on the bridge of his nose beginning in January, 1983.
Cutaneous lesion due to Leishmania mexicana on ear lobe
of patient B. Photo courtesy Dr. T. Gustafson.
Cutaneous lesion due to Leishmania mexicana on thigh of
patient C. Photo courtesy of Dr. T. Gustafson.
Figure 2-4. Cutaneous lesion due to Leishmania mexicana on cheek of
patient D. Photo courtesy of Dr. T. Gustafson.
When the lesion persisted the boy was seen by a physician who biopsied
it before a positive diagnosis could be rendered. Tissue samples and
saline aspirates taken from the biopsied lesion contained no
parasites. The boy had not traveled outside the USA, nor outside the
region during the previous year. The family had three pet dogs.
Serologic tests performed for each patient included indirect
fluorescent antibody test (IFA) and dot enzyme-linked immunosorbent
assays (ELISA) at Walter Reed Army Institute of Research, Wash., DC,
and fluorescent immunosorbent assays (FIAX) at Oklahoma State
University, Stillwater, Oklahoma (Gustafson et al., 1984). These tests,
performed between three and six months after onset, showed that all
four confirmed patients had detectable antibody titers to Leishmania.
The sister of Patient C and the mother of Patient D had detectable
titers by at least two serologic tests. One dog belonging to Patient
C and three dogs belonging to Patient D had detectable Leishmania
titers by at least two tests. However, all four dogs had antibody
titers to Trypanosoma cruzi, and the Leishmania titers may represent
cross-reactions (Gustafson et al., 1984).
Description of Study Sites
Eight study sites were selected based on previous collection
records of Lutzomyia from south central Texas (Eads et al., 1965; Easton,
et al., 1968; Young, 1972; Eads, 1978; Endris, 1982; Perkins, 1982) and
on their association with confirmed Leishmania patients (Fig. 2-1).
Three of the sites, A, B, and D, were close to the home environs of
patient A and he had camped at or near the sites three months prior to
onset of symptoms. Lutzomyia species were also known to occur at site
A (Young, 1972; Endris, 1982). Site B was selected on the basis of
1971 light trap collection records of 237 females from a single light
trap (Perkins, 1982). The remaining sites E, F, G, and H were the
actual home environs of confirmed leishmaniasis patients.
Surveys for potential sand fly vectors were conducted at all sites
during the course of the study. Emphasis was placed on qualitative
rather than quantitative sampling; no attempt was made to estimate
sand fly population densities.
Site A. Garner State Park, Concan, Uvalde County, Texas (4-28
June, 1982; 19-30 September, 1983; Fig. 2-5 and 2-6). This state park
is located 50 km north of Uvalde in the Frio River valley. It is
situated on the southern edge of the Edwards Plateau, on the Balcones
Escarpment, which separates the plateau from the Nueces plains to the
south. The habitat is characterized by grassy meadows dotted with
cedar (Juniperus sp.), live oak (Quercus virginiana), acacia (Acacia
sp.), pecan (Carya illinoensis), mesquite (Prosopis juliflora), and
elm (Ulmus crassifolia), with bald cypress (Cupressus arizonica)
bordering the Frio river. The park is surrounded by steep rocky
hills, the slopes of which are covered with cedar, wild cherry (Prunus
serotina ), persimmon (Diospyros texana), madrone (Arbutus texana),
and other Hill Country shrubs. Potential mammalian hosts for sand
flies include white tail deer (Odocoileus virginianus), coyote (Canis
latrans), fox (Vulpes sp.), bobcat (Lynx rufus), racoon (Procyon
lotor), porcupine (Erethizon dorsatum), skunk (Mephitis sp.), opossum
(Didelphus marsupialis), armadillo (Dasypus novemcinctus), jack rabbit
(Lepus californica), cottontail rabbit (Sylvilagus floridanus), and a
variety of rodents (Texas Parks and Wildlife Department, 1982a; Raisz,
1954; Kuchler, 1975).
The nearest weather station to Garner State Park is 50 km south at
Uvalde where the climate is fairly representative of the area
surrounding the park, characterized by hot, humid summers and
pleasantly mild winters (US Department of Commerce, 1965). Annual
precipitation averaged 63.1 cm between 1951 and 1960, with a range of
23.3 cm to 98.3 cm. The wettest periods of the year are May-June and
September-October. The hottest part of the year is July-August, with
extreme temperatures reaching a high of 430C, and a low around -150C,
reached in January and February. Relative humidity at noon, central
standard time, averages between 58% in January to 48% in July. Daily
fluctuations in RH occur especially during the warmer months, with a
gradual decrease in the afternoon until sunset, then a rapid increase
after dark (US Department of Commerce, 1965). Average daily maximum
and minimum temperatures for June 1982 were 350C and 220C,
respectively, and precipitation 14.4 cm (US Department of Commerce,
1982a). Average daily maximum and minimum temperatures for September
1983 were 340C and 210C, respectively, with 7.9 cm precipitation (US
Department of Commerce, 1983a).
Site B. Rio Frio, Real County, Texas (4-20 June, 1982). This
site is located approximately 20 km north of Site A, at the edge of a
small gorge carved by the Frio river. The habitat and climate are
very similar to that found at Garner State Park.
Site C. Seminole Canyon State Park, Val Verde County, Texas (10-
11 June, 1982; Fig. 2-7). This state historical park is located 75 km
west of Del Rio, Texas, a short distance downstream from the
confluence of the Rio Grande and Pecos Rivers. It is noted for its
rugged terrain, sparse vegetation and deep canyons. Situated in the
arid Pecos Shrub Savanna, its flora and fauna are derived from
elements of the Texas Hill Country, Tamaulipas Thorn Scrubland, and
the Chihuahua Desert (Texas Parks and Wildlife, 1982b). It is
considered a true desert habitat, receiving 25 cm or less
precipitation annually. The nearest weather station, Amistad Dam,
recorded average maximun and minimum temperatures for June 1982 of
360C and 230C, respectively, with daily temperatures frequently rising
above 38C during the summer months (US Department of Commerce,
1982a). Relative humidity was not measured, but probably averaged
below 30% during the summer. Mammalian inhabitants at the park are
limited to coyote, jack rabbit, armadillo, and small rodents.
Site D. Fawcett Boy Scout Camp, Barksdale, Edwards County, Texas
(11 and 22 June, 1982; Fig. 2-8). This site is located 7 km north of
Barksdale at a bend in the Nueces river. Also at the southern edge of
the Edwards plateau, the habitat resembles Garner State Park in most
features, but lacks steep, rocky hills. The nearest weather station
is about 20 km south at Camp Wood (US Department of Commerce, 1965).
Annual precipitation for the ten-year period of 1951-1960 averaged
59.8 cm with a range of 22.3 cm to 106.1 cm. Average daily maximum
and minimum temperatures during June 1982 were 330C and 210C,
respectively (US Department of Commerce, 1982a).
Site E. Myer's Farm, D'Hanis, Medina County, Texas (19-30
September, 1983). This farm, the home of patient D, is in a small
rural community 75 km west of San Antonio. The surrounding area is
characterized by low rolling farmland with live oak, pecan, and
mesquite trees bordering the fields and streams (Fig. 2-9).
Uncultivated areas are overgrown with cactus, mesquite and thorny
shrubs. One such area, about 50 m west of the farm house, contains a
number of woodrat (Neotoma) nests and armadillo burrows. Roughly 15 m
south of the house is a large equipment shed, behind which (to the
west) is the kennel in which are kept the family's ten hunting dogs
(Fig. 2-10). Between the house and the kennel are a large live oak
tree and a wood pile (Fig. 2-11). Approximately 80 m west of the
house are the hog pens (Fig. 2-12). Approximately 100 m behind (west
of) the house is a dry creek bed lined with mesquite trees, and beyond
that, a grove of live oak trees. The closest weather station is in
Hondo, 20 km to the east. The climate at this station is
representative of Medina County, with hot humid summers and pleasantly
mild, dry winters (US Department of Commerce, 1983b). Average annual
rainfall is 71.2 cm with heaviest rainfall received during April-June
and September-October. Mean length of the warm season is 263 days.
Mean dates of last occurrence of freezing temperatures in spring, and
first occurence in the fall are March 6 and Nov. 24, respectively.
Humidity at noon central standard time averages 58% in January, 56% in
April, 48% in July, and 55% in October. Average daily maximum and
minimum temperatures for September, 1983, were 320C and 260C,
ranged between 430C
minima of 270C and
Site F. Romer
September, 1983; Fi
grandparents of Pat
it is very similar
Site G. Romer
13.0 cm precipitation. Temperatures for the year
and -160C with ten-year average daily maxima and
130C (US Department of Commerce, 1983b).
Ranch, Devine, Medina County, Texas (19-30
g. 2-13). This is the home of the paternal
ient B. Approximately 10 km south east of Devine,
in general features to Site E in D'Hanis.
Residence, northwest San Antonio, Bexar County,
Texas (22 September, 1983; Fig. 2-14). This is the home of Patient B
Habitat typical of that found at Garner State Park, Concan,
Uvalde County, Texas.
Habitat at Garner Stake Park, Concan, Uvalde County, Texas,
showing public latrine resting station.
Habitat typical of that found at
Val Verde County, Texas.
Seminole Canyon State Park,
Habitat at Fawcett Boy Scout Camp, Barksdale, Edwards County,
Texas, showing open latrine resting station.
Farmland surrounding the home of patient D in
community of D'Hanis, Medina County, Texas.
Hunting dogs in a kennel behind the
D'Hanis, Medina County, Texas.
home of patient D in
~,. *` 'h
Wood pile located behind the home of patient D in D'Hanis,
Medina County, Texas. Six female and one male Lutzomyia
diabolica were collected in the CO2-baited CDC light trap
hanging in the foreground.
Hog pen located about 80 m west of the home of patient D
in D'Hanis, Medina County, Texas. Rock pile in foreground
offered many potential resting places for sand flies.
Romer ranch, near Devine, Medina County,
the paternal grandparents of patient B.
Texas, home of
Neighborhood of patient B in northwest San Antonio, Bexar
ir~ m _.: o, ;.: ..
~ ri .. .. '" -.-" .,.. "
and is in a typical suburban neighborhood on the northwest limits of
San Antonio. About three blocks south of the residence are some open
fields, heavily overgrown with weeds and small shrubs. An equal
distance to the north is a small creek basin lined with a variety of
large shade trees and shrubs.
Site H. Kelly Residence, San Antonio, Bexar County, Texas (23 and
30 September, 1983; Fig. 2-15). Similar to site G, this site is in a
suburban neighborhood in southeast San Antonio. Three large pecan
trees shade the backyard. Adjoining the property to the southeast is
a large vacant lot, overgrown with tall weeds.
San Antonio climate is about the same as that of Medina County
(US Department of Commerce, 1982b) with hot humid summers and
pleasantly mild, dry winters. Average annual precipitation is
74.6 cm. Average daily maximum and minimum temperatures during September
1983 were 320C and 200C, respectively, with an overall mean of 260C.
Precipitation during September, 1983, measured 13 cm (US Department
of Commerce, 1983a).
Materials and Methods
Determination of Sand Fly Fauna
Several useful field methods for sampling sand flies have been
used with varying degrees of success and are reviewed by Young (1979)
and Killick-Kendrick (1981a). The diversity of potential sand fly
habitats at each study site dictated the selection of methods that
would sample the broadest cross section. As far as practicable, all
representative habitats within a 200 m radius of each site were
Figure 2-15. Neighborhood of patient
A in southeast San Antonio, Bexar
surveyed to determine species diversity and relative abundance of sand
flies. Collection methods employed included the following:
Resting collections. An extensive search was made at each site
for diurnal and nocturnal resting places of adult sand flies.
Particular attention was directed to dark protected cavities in rocks,
trees, buildings and other man-made structures, as well as in animal
burrows, ground nests and brush piles where sand flies might seek
shelter. A flashlight was used to enhance visibility. Live specimens
were captured with a simple tube aspirator and placed in a 120 ml
feeding/rearing chamber (Endris et al., 1982).
Biting collections. Biting collections (Young, 1979) were made
at night, usually between 2100 and 2400 hrs. The collector sat in an
open area under an incandescent light or Coleman lantern and captured
sand flies with an aspirator as they attempted to feed.
Light traps. Battery powered or CDC miniature light traps (Sudia
and Chamberlain, 1962) were secured to tree branches about 2 m above
the ground at edges of clearings and near human habitations. Care was
taken to insure their placement was away from competing light. In
some instances they were placed adjacent to or directly over suspected
resting sites such as wood or rock piles (Fig. 2-11).
A New Jersey mechanical light trap (Mulhern, 1942), powered by
110-volt house current, was placed in the back yard, adjacent to the
dog kennel at the D'Hanis site. The trap will be operated there 2-4
nights per week for the next year.
A Shannon trap (Shannon, 1939; Fig. 2-16), made from white cotton
bedsheets, was erected in open areas near human and animal dwellings.
This apparatus does not actually trap insects, but the lantern or
Figure 2-16. Shannon trap erected behind the home of patient D, D'Hanis,
Medina County, Texas.
kte*Q r-lii' r
incandescent light inside provides a light source, enabling the
collector to aspirate flies which land on the illuminated cloth.
Flies are attracted to the light, the collector, or a combination or
both (Young, 1979).
Baited traps. Pieces of dry ice wrapped in newspaper and
suspended next to the CDC light traps provided CO2 as an adjunct
attractant for hematophagus insects (Fig. 2-11). A caged hamster,
suspended in a similar manner, also served as an attractant.
A Disney trap (Disney, 1966) baited with a hamster or wild cotton
rat (Sigmodon sp.) was also used. This trap consists of a shallow
tray filled with mineral oil and a caged animal supported on slats
just above the surface of the oil. After feeding, blood engorged sand
flies do not fly away, but hop to the "ground" to rest, becoming
entrapped in the mineral oil.
Sticky traps. Sheets of card stock, coated on one side with oil,
were placed directly in front of or over burrow entrances and crevices
in rocks and trees to catch sand flies emerging from such places at
night (BUtticker, 1979).
Processing and Maintenance of Wild-caught Sand Flies
and Recovery of Eggs
Wild-caught female sand flies showing evidence of a blood meal
(abdomen distended and dark-red or black in color) were transferred to
individual 7-dram oviposition vials fitted with screen lids (Endris et
al., 1982). Water was added to the plaster of Paris in the bottom of
each vial to maintain a high relative humidity, and a drop of Karo
syrup in water (1:1) was placed on the screen lid to serve as an
energy source. Unfed flies were offered a blood meal, through the
screen lid of the holding vial (Fig. 3-1, p. 98, Chap. 3) on a human
forearm, a hamster or a lizard. Those that fed immediately were
transferred to individual oviposition vials, as previously described,
and held until they oviposited and/or died. Females that did not
accept the initial offer of a blood meal were placed in a holding cage
(Endris et al., 1982) with males and offered a blood meal each
succeeding day until they either fed or died. Those that fed were
retained in oviposition vials until they oviposited and/or died. Dead
females were removed and, if they had deposited eggs in the vials, the
screen lids were replaced with solid plastic lids that had been
perforated with a dissecting needle to permit limited gas exchange.
During June 1982, the field laboratory consisted of a permanent,
screened camping shelter. To protect the captured flies from the heat
of the Texas summer, their containers were kept in a polystyrene
cooler. By placing ice in the cooler, removing or replacing the lid
as necessary, the temperature within the chambers and vials could be
maintained at or near 270C. During the September, 1983, trip the
field laboratory was in an air-conditioned motel room where the
temperature was maintained at 230C.
Preliminary identification of live, wild-caught specimens was
performed macroscopically. At death they were more reliably
identified by microscopic examination and comparison with appropriate
taxonomic keys (Young and Perkins, 1984). Field identifications were
later confirmed as necessary in the laboratory by Dr. D. G. Young,
after consulting appropriate keys and available type materials.
Specimens were prepared and cleared for permanent slide mounts
according to the technique of Young (1979).
Post mortem dissections of female sand flies were performed as
soon after death as possible (usually within 12 hours) according to
the technique of Johnson et al. (1963). Data recorded for each
specimen included information regarding collection, sex,
identification, blood meals, oviposition, longevity, condition of
accessory glands, and presence or absence of natural parasites.
Species of Sand Flies Present
An estimated 2400 sand fly specimens representing five species of
Lutzomyia, including one new species, were collected and eight new
county records established during the two years of the study (Table
2-1). Sand flies were taken from five of eight case sites; none were
collected at sites F, G, and H (San Antonio and Devine).
Of the sampling methods used, only the Disney trap and the oil
traps failed to collect sand flies. Approximately 2000 flies, mostly
females, were taken in resting collections at Garner State Park and
Fawcett Boy Scout Camp. Other trapping methods also yielded mostly
females (Table 2-2). The ratio of male to female Lu. diabolica in
resting, biting and light trap collections ranged from a high of about
6:10 in resting collections at Fawcett Boy Scout Camp, to around 4:10
in resting collections at Garner State Park. More than 350 female and
no male Lu. diabolica were taken in biting collections at Garner State
Park. This method was not successful at other sites. At locations
where flies were not taken resting, CDC light traps were most
e c 0,
*r- *i- U) C
4-)I Sr- 5-
l 4-l- 0
(u- 3 (o E
4- S- -0
I C O--
0 0 0 ln- (nl
C C 4- C *r- 3
*r- *- r C- 'o o
l 4 *- l- E s-
5. r- .-- E (z -0
s. u E'a.
4- 0) *-
Qn &*E -
10 0 0
CS- 4- S- S-
I ea 1 4-) -P 4j
S4 O 4- 0 4-
S- r I I- EC
- S.,- C --.C Cr
(D >l C 0 01- 0- l 0 '-
S r- -E QC- CO /-)
I 3 C O -
CE C 4- C *- 3
*- *r- JC *- 0
4-) S- 01 C 'e 5-
(V 4-J E 5-
ai to C 0 -
5s.- E-' S o .0
L 0 o
l 4- *r-
p S- C 4 0 -
S4--* V-C..) 0 r
0 CO CO
0 ) 0
CIC 4-' C4-P
s-i-- s- S --
C: 4-. -Z
< S.- S-
C /)0 CL
(C. E C M
W CO 4-P
oo C-) oo
0 3 >, E
'a O0 r
Ll.CO M u
L aJ S.-
ii (L ..
C C 4-D 0C O -
0) C a a .C 0 5.. .C 0
C O -- 00- 00 0- 0)0-- O -- ,,
en &" Ou c- .- *-x --* --- o
SC C .C CD C C-
-C O 0 e- 0 0 0 0 .
Oi 1 ---- O ,--0 0
*I- 4 4
CU 4- In -)
0) 0- .. -a
+-) 4- 4- -C +-C
.. I- 4 0 O r S.
.C 0 -.C C 0 S .C
4-4- I I I I O ) r-41. 0 > a
-J 0 -' r- -- N -- -- C .C
E L- C 0 C 4.-
o r3 0-\ .- O -0 4O r0 0>
.- 2 0 I I I I 0 0 < r-
4- '- C
U) CN Un 4.-)
C CN C 0.
o 0 -. 00
*r- E 4-
4 0 + rC -E V 0a
U O Q. -- ) 3 C -o 4- r-
a) -C s- 1 a r-- to *- ad Lo u B
- 4- I-- 4- I I I I C CO C -4 CO 4->
r-J o .... x r O
0C I I I I CO a) ,,)0
O S 4-> O O LO O E a
U0 0 I ) 04
U 01 u I- 00
m c -u a. ) a. )
4-3 0O 4-) u =h 3L Q
< a).') C M- 0
U) 0 5.. -- .-. --.
C 0 4- a --) .. C '-- -* C .--
4i 4- 4- 4-1 4- 4-
C -CL -r .-C= kO 01 ) -C *-*-' _r = 1 ko
*r- O 4-) 4-) CM an r-r CD- o Co cm C (M C cm CM f
0 S ..- C-- ~-0 C:) C CD CD -
(0 <.2 M r~ CM O. O. O^ ON CM X
*-0 --4 0 r -4 O O -4 C m I-
- I .0 -' -
D 0= -- -- -C X "
*- 0*~ O 0 44 0 -- L-
Co -. a) .C" 0-.C" -*-E.C -. -- o 0
CE Z 0 C 4- O 0 <-) O 00 O 00) 0-- O 3
N C 4- C C C O uj x
4+J- *r 0C 0 0N 0 -. 0 0 C 1-C
CO -0 0* LOCV 0 0
j M M-- a)
0I- I f a
E .- =- I --- =3 -l
* o *0 CN mD r r- mD m I- O mC O m) CM ro w
0 4-a ) S- r r CL .
a 0 *- 4 -N .C- 0srC C4-4 C CD -% O E "0E
cU E C o CO 0 c N -- o o *CN -z C-) a 0V)
r D C -
4- a- ( -
u 4-) x 4-' E
0 0w b o+ b oC b o+ 1b o0 (aO L-
3EX 0 CO U-
* (U 0)-
CM Ea u S-
I S- S a)
v, CD .LL
S* 0 II II IIL
Lutzomyia diabolica. Lu. diabolica (Fig. 2-17), the only
anthropophilic sand fly found, was collected at all positive sites and
accounted for 99% of the total catch (Table 2-1). They were collected
in greatest abundance at Garner State Park where they were captured
while resting on the tile walls of the public latrines (Fig. 2-6).
There were a total of 12 latrines at the park, 8 of which were
situated in the upland meadow areas and 4 along the Frio River flood
plain, shaded by large pecan, oak, and cypress trees. Up to 150 Lu.
diabolica per day (nights and mornings) were taken in resting
collections from the 8 latrines in the more elevated portions of the
park. Many were engorged with fresh blood, possibly from the
unprotected parts of unsuspecting campers, or from deer and jack
rabbits that frequent the surrounding meadows by night. No sand flies
were collected from the latrines near the river. Although a few sand
flies could usually be found in positive latrines at all hours of the
day, they typically appeared about one hour after sunset (2200 hrs in
June, 2100 hrs in September) and remained until mid morning (about
1000 hrs), their activity coinciding with the daily bathing ritual of
the campers. They were observed resting on the interior tile walls
between the level of the floor and a height of about 2.5 m and seemed
to prefer dark humid corners, such as in shower stalls and under sink
counters. Their resting attitude was almost always vertical, with
head pointing up. They were frequently observed hopping deliberately
toward the collector, presumably seeking a blood meal. If by chance
the lights in a latrine had not been turned on, the catch was greatly
reduced. Greatest numbers appeared on hot, humid nights (270C and 80%
or higher RH) with little or no air movement. On such nights sand flies
Figure 2-17. Lutzomyia diabolica female (approx. magn. x 10).
were also seen resting on the exterior walls of the latrines under the
outside lights and around the entrance. The number of flies collected
decreased dramatically with increased wind velocity. At Fawcett Boy
Scout Camp, sand flies were also taken in resting collections in a
lighted, open-air latrine (Fig. 2-8).
An extensive search for natural diurnal and nocturnal resting
sites of Lu. diabolica in rock crevices, under bark of trees, in
animal nests and burrows, in leaf litter, and other potential hiding
places was fruitless at all case sites.
Female Lu. diabolica were often taken in biting collections during
the evening hours at Garner State Park as they came in search of a
blood meal. Routine biting collections netted from one to about
twenty female sand flies per hour. One biting collection, however,
was worthy of special note: The night of 9 June, 1982, was hot and
humid (270C and 80-90% RH) and air movement was about nil. The author
and his son had returned to the field laboratory (a screen tent) about
2330 hrs after searching the latrines for resting sand flies. They
sat under an incandescent light at a table inside the tent and began
to count and feed the evening's catch. Shortly the author's son began
to complain of being bitten by sand flies, and five were collected
that were biting him on the legs. Lu. diabolica females flew
unobstructed through the screen sides of the tent in waves of ten or
twenty to feed, biting the occupants on the face, neck, hands, ankles
and other exposed skin. These foraging sorties were of about five
minutes' duration, with a brief respite interval between. They
continued for several hours and about 150 female Lu. diabolica were
collected before 0100 hrs, when the author finally retired for the
night. Even with the light turned out, the flies continued their
attack, but in lesser numbers.
On several occasions, foraging females were seen flying through
the standard mesh screen of the permanent camping shelter (field
laboratory) to attack the author and his eight-year-old son. They
seemed to exhibit a preference for the son, biting him even after the
lights had been turned off. Although most of this biting activity
occurred at night, one female Lu. diabolica bit him at about 1400
hours in the heat of the day and in full sunlight (temperature about
380C, RH about 60%, air calm).
Lu. diabolica were collected in unbaited CDC light traps at Garner
State Park, Rio Frio, and Seminole Canyon State Park in June 1982.
They were also collected in CO2 baited CDC light traps (with light on)
and in a New Jersey light trap at the site in D'Hanis in September,
1983. Unbaited CDC light traps (light only) were unsuccessful at the
One female Lu. diabolica was collected in a Shannon trap at Garner
State Park. The trap was set up near the camping shelter about an
hour before dark (1800 hrs), but was only operational for about three
hours due to inclement weather. Other attempts to attract sand flies
to a Shannon trap at D'Hanis and Devine failed.
Disney traps set out near rodent burrows at Garner State Park
failed to collect sand flies as did oil traps set out in similar
places at D'Hanis and Devine.
Lutzomyia anthophora. Lutzomyia anthophora were taken with Lu.
diabolica in resting collections in latrines at Garner State Park and
Fawcett Boy Scout Camp. They were also taken from armadillo (Dasypus
novemcinctus) burrows at the former site.
One male Lu. anthophora was taken from an active, carefully
dismantled woodrat (Neotoma) nest located approximately 75 m east of
the farm house at the D'Hanis site (Fig. 2-18). Four other Lu.
anthophora (males and females) were seen in the same nest, but escaped
Lutzomyia texana. Lutzomyia texana, a rather large sand fly that
inhabits mammal burrows, was taken in resting collections from
latrines at Garner State Park and Fawcett Boy Scout Camp. They were
also collected from the entrances of armadillo burrows at the former
site and in CO2 baited CDC light traps and in an unbaited New Jersey
light trap at the farm in D'Hanis. They were not found in armadillo
burrows at the latter site. Efforts to feed Lu. texana on a human, a
hamster and a lizard were unsuccessful.
Lutzomyia vexator. This species, which feeds on cold-blooded
vertebrates, was collected on two occasions while resting on latrine
walls at Garner State Park.
Lutzomyia new species. In September, 1983, a single female of an
undescribed species was taken in a resting collection from a latrine
at Garner State Park. Efforts to feed the fly on human, hamster and
amphibian (toad) blood were futile and she died without depositing
eggs. The fly was prepared and examined by Dr. D. G. Young who stated
that it belongs in the Cruciata group, but was unlike any other
species collected in Texas.
Processing and Maintenance of Wild-caught Sand Flies
and Recovery of Eggs
An estimated 1925 female Lu. diabolica were segregated and
maintained in individual or group oviposition containers and 7940 ova
p~ c~~- :~fj;r ~ *L.
.4;,,-. *, I *'4 ~ C% 'V
Dismantled woodrat (Neotoma) nest near the home of patient
D in D'Hanis, Medina County, Texas, where specimens of
Lutzomyia anthophora were found.
were recovered for laboratory colony stock. Six hundred and one
females were dissected post mortem and examined for retention of eggs,
condition of accessory glands and natural parasite infections.
Table 2-3 summarizes fecundity for 360 wild-caught females (263
from the first trip and 97 from the second). Regardless of efforts to
perform dissections within 24 hrs after death, many flies were in such
poor condition that dissection provided little usable information.
These and 54 females that died from overheating in a locked car in
June 1982 were excluded from the table. The mean number of eggs
deposited and the gross number of eggs produced (the number deposited
plus the number retained) remained fairly constant from one year to
the next. From 31 to 37% of the females retained some or all of their
eggs. Females neither depositing nor retaining eggs (in other words
maturing none) ranged from 3% in June 1982 to 10% in September 1983.
No evidence of autogenous behavior was observed. Females that were
denied a blood meal neither laid nor developed mature eggs.
Significant differences were noted between year groups in
preoviposition interval, postcapture longevity and postoviposition
longevity, all three being longer for flies captured in the fall
(Table 2-4). A smaller percentage of females survived oviposition in
the fall than in the spring.
Accessory Glands and Parity
Sixty-four wild-caught female Lu. diabolica, which had not been
offered a blood meal, were dissected to study the paired accessory
glands (Fig. 2-19). The glands of only 36 flies were clearly seen.
Of these 36, only three showed evidence of having had a blood meal;
co .O CMj
CO ;) -) ,-- -)
0 I r-r S ->I
o ) o (o *=- ) c C ) 0 i) O
*- /1 *- *- *r- *r- a) 4- 0 "0 C4- 0
,T >b >i( 1 -o oJ o7 +O oE 4- 0
(B en 0c 3 (. CT i- o L. o O 0
-JI -i +.2 Q*' (U cr" s 2 i l- a.1- ..
LD mL" r-
cr CM r)
Table 2-4. Summary of preoviposition intervals, postcapture and post-
oviposition longevities in wild-caught Lutzomyia diabolica
females from south central Texas (based on a sample of 100
females per year group) (4-28 June, 1982; 19-30 Sept., 1983)
June 1982 September 1983
Category n/100 x 1 SD (range) n/100 x 1 SD (range)
Captured w/ 31/100 4.5 2.9 (1-10) 46/100 8.6 3.7 (1-16)
Captured w/o 69/100 4.7 2.1 (1-11) 54/100 8.5 3.8 (1-16)
Overall 5.8 3.0 (1-19) 9.7 4.2 (1-19)
Laying Some 56/100 6.5 6.5 (1-16) 75/100 9.6 3.79 (3-19)
Maturing No 3/100 5.6 2.1 (3-9) 10/100 10.4 5.4 (1-19)
Taking One 56/100 2.2 1.7 (1-8) 29/100 3.2 3.2 (1-11)
Taking Two 0/100 4/100 4.7 2.3 (2-6)
all three had either mature or developing ova and the accessory glands
were full of dark granules. Of the remaining 33 females, 17 had
developing ova (most were about half developed) and all of these had
either full or partially full accessory glands. Sixteen had unde-
veloped ovaries,and of these, five had no granules in the accessory
glands; one had a very few granules; nine had accessory glands half
full of granules; and two had completely full accessory glands.
Observations on the condition of the accessory glands (granule
formation) in another 94 dissected wild-caught females are presented
in Fig. 2-20. All of these females were blood-fed at the time of
capture or were given a blood meal immediately after capture. These
were compared with dissections of laboratory-reared, fed and unfed
females of known ages. In general, granule formation in accessory
glands of wild-caught sand flies paralleled ovarian development, and
when the ova were mature, the accessory glands were dark and full of
granules. Partial or complete emptying of the accessory glands often
occurred as eggs were deposited, but many females that deposited most
or all of their eggs maintained full accessory glands. Blood-fed
laboratory-reared flies showed basically the same trend as blood-fed
wild-caught flies. Unfed laboratory-reared flies either did not
develop granules in the accessory glands or, if they did, they
developed them within 24 hrs, but to a lesser degree, even after six
days, than seen in blood-fed females.
Natural Parasite Infections
The incidence of natural parasite infections found in 341 female
sand flies taken from south central Texas, dissected during the two
years of the study, is presented in Table 2-5.
Paired accessory glands of a gravid female Lutzomyia
diabolica showing dense granular material (magn. x 160).
2 l | | |
0l ) 0 ra r
6 o 1 -
x w 0 D
in o 0
- & e m 0 A
L -- 2 S
E2 -0 C 4e
S5 5 0
oco o -
w a2 U)'O
ai~ C, Q
a g~g LL.
Incidence of natural parasite
Lutzomyia diabolica collected
(4-28 June, 1982; 19-30 Sept.,
infections in 341 female
in south central Texas
June 1982 (N = 243)1
1983 (N = 98)2
1. Specimens collected primarily from Garner State Park, Uvalde County,
Texas, with small numbers collected from Fawcett Boy Scout Camp, Edwards
County, Texas, and Rio Frio, Real County, Texas.
2. All specimens collected from Garner State Park, Uvalde County, Texas,
with the exception of six specimens collected from D'Hanis, Medina
Flagellates. Small, rounded, highly motile flagellates were
found swimming free in the hemocoel of 27 to 60% of the specimens
dissected. They were most often seen in the hemocoel of the anterior
thoracic region and immediately posterior to the head. They numbered
from one to several, but rarely more than 25. Efforts to photograph
these tiny parasites were unsuccessful.
Two females collected in the fall of 1983 contained what appeared
to be small thin flagellates in their midguts. The organisms did not
move and appeared to be dead. They were stained with Giemsa and
mounted on microslides with Euparal, but were unidentifiable.
An infection of unknown epimastigote-like flagellates was
discovered in the midgut of one of the females dissected in the spring
of 1982 (Fig. 2-21). Several hundred of these promastigotes were
observed freely swimming in the lumen of the gut; some attached to the
midgut epithelium by their flagella; others formed rossettes with
their flagella directed toward the center much like L. mexicana
parasites seen in laboratory infections and in cultures in vitro.
Their general appearance, however, was not like Leishmania. The
infected fly was unfed when captured and was given a blood meal on the
author's forearm. Upon dissection (six days later), the blood meal
remnant was evident in the midgut. Midgut contents containing the
flagellates were inoculated subcutaneously into the foot pad of a
hamster. No lesions or other signs of infection developed in either
the author or in the hamster.
Gregarines. Aseptate gregarines, possibly Ascocystis chagasi
(Adler and Mayrink), were found in 7-9% of dissected females. Up to
five gamonts were observed in the abdominal cavity of some infected
flies (Fig. 2-22). These could be seen slowly changing shape and
moving in amoeboid fashion. When present, one to three gametocysts
could be seen attached to the accessory glands (Fig. 2-23). The
eliptical oocysts were most frequently seen spilling from the
accessory glands or from ruptured, mature gametocysts (Fig. 2-24).
Oocysts were seen glued by the accessory gland material to the chorion
of eggs deposited by gregarine infected females.
Microsporidians. Massive microsporidian infections were
discovered in the hemocoels of 3-8% of the dissected females (Fig.
2-25). Germination of the spores was stimulated in vitro by adding a
few drops of 0.2m KC1 (pH 9) to the dissecting medium (insect Ringers
solution) (Fig. 2-26). Ungerminated spores washed onto the larval
medium were fed to uninfected larvae. No infections were acquired.
Mites. Mites were found attached to the venter and dorsum of the
abdomen of 14 females collected in the spring of 1982. None were
found on females collected in the fall of 1983. The mites were
tentatively identified as Eustigmaeus sp. (Berlese) (Family
Stigmaeidae). Scanning electron micrographs of one mite specimen are
presented as Figures 2-27 and 2-28.
Other. Fungal and bacterial infections were commonly found in
dissections of wild-caught Lu. diabolica.
Discussion and Conclusions
Determination of Sand Fly Fauna
Many methods have been used for sampling sand fly populations.
Some were devised for sampling particular habitats (e.g., resting
Unidentified flagellates in the midgut of a Lutzomyia
diabolica female collected at Garner State Park, Uvalde
County, Texas, June 1982 (approx. magn. 2100).
of aseptate gregarine in the abdominal cavity of a
Lutzomyia diabolica collected at Garner State Park,
County, Texas, June 1982 (approx. magn. x 850).
Gametocyst of an aseptate gregarine (center) attached to the
accessory gland of a female Lutzomyia diabolica collected at
Garner State Park, Uvalde County, Texas, June 1982 (approx.
magn. x 850).
Gregarine oocysts spilling from the accessory glands of a
female Lutzomyia diabolica collected at Garner State Park,
Uvalde County, Texas, June 1982 (approx. magn. x 2100).
Microsporidian spores in the hemocoel of a female Lutzomyia
diabolica collected at Garner State Park, Uvalde County,
Texas, June 1982 (approx. magn. x 2100).
Microsporidian spore germinated in vitro by addition of
0.2 M KC1 (pH 9) to dissecting medium(insect Ringer's
solution, pH 7.2) (approx. magn. x 2640).
Scanning electron micrograph of mite (Eustigmaeus sp.) found
on the abdomen of a female Lutzomyia diabolica collected at
Garner State Park, Uvalde County, Texas, June 1982 (dorsal
aspect; magn. x 1704).
;,, : ,i"
Scanning electron micrograph of mite (Eustigmaeus sp.) found
on the abdomen of a female Lutzomyia diabolica collected at
Garner State Park, Uvalde County, Texas. June 1982 (ventral
aspect; magn. x 1700).
collections, sticky traps, Disney traps) and are of little value
elsewhere; others have broader utility (light traps) and can be used
regardless of the availability of suitable habitat. From the data in
Table 2-1 it appears that the latrines at Garner State Park and at
Fawcett Boy Scout Camp were universally attractive structures. All
species taken at these two sites were taken in resting collections in
the latrines. In other localities, where resting sites were not
discovered, even after extensive searching, light traps proved most
effective. Biting collections were effective as a selective means of
sampling anthropophilic species at Garner State Park, and probably
would have been useful at other sites as well. Baiting of CDC light
traps with dry ice also selected for man biters and was the only way
Lu. diabolica were collected at the D'Hanis site.
Due to the success of collections in latrine resting stations at
Garner State Park and Fawcett Boy Scout Camp, other methods such as
light trapping and bait trapping were not routinely used.
Most sampling methods collected predominantly females, which is
consistent with findings of other authors (Young, 1972; Endris, 1982).
No males were taken in biting collections, although some would be
expected to mate with females on the host as she feeds. The figures in
Table 2-2 should not be interpreted to represent the true sex ratios
in the natural population. Most trapping methods, especially those
using light or bait attractants, select for females. This is
especially true if the female's movement is part of a hunting strategy
(Killick-Kendrick and Rioux, 1981). Chaniotis et al. (1972) stated
that the true sex ratio could only be properly assessed by studying
populations from each habitat using a variety of sampling methods
(especially those that do not select one sex over the other).
Lutzomyia diabolica. This species was discovered in Uvalde, Texas,
in 1915 and was said to attack man freely (Parman, 1919). It was
subsequently collected by other workers in areas limited almost
exclusively to the south central part of the state (Young and Perkins,
1984; Fig. 2-29). Disney (1968) held that Lu. diabolica was
conspecific with Lu. cruciate, a species widely distributed through
Central America, and a suspected vector of leishmaniasis in Belize
(Williams, 1966a, 1966b). Young and Perkins (1984) showed that Lu.
diabolica from Texas represents a distinct species, separate from Lu.
Parman (1919) found Lu. diabolica hiding in dark places during the
day, one or two specimens to a place. He offered few clues as to
where these "dark hiding places" were, except to say there was evidence
that the breeding places are in neglected poultry houses, since the
flies were observed in abundance around such places in the late
twilight hours. Lindquist (1936) captured male and female Lu.
diabolica on walls and curtains in lighted rooms on first and second
floors of dwellings. Endris (1982) collected Lu. diabolica resting in
latrines at Garner State Park. All resting collections of Lu.
diabolica have been made in or on human dwellings or outbuildings. To
date, a "natural" resting habitat of this species has not been found,
indicating that it may be a peridomestic sand fly. This underscores
the need for future searches to locate natural resting places and
pinpoint breeding sites. Resting Lu. diabolica were collected in
latrines only in the upland portions of Garner State Park, an
S= autochthonous human case of
E= known distribution of
Known geographical distribution of the sand fly Lutzomyia
diabolica in Texas.
observation consistent with Parman's statement (1919) that the sand
fly was found in Uvalde in the more elevated parts of the city.
Perhaps this is a clue as to where other resting sites are to be
Parman (1919) reported that the earliest authentic record of
appearance of Lu. diabolica in Uvalde was 3 September and the latest
was 24 November. Lindquist (1936) reported collections between 3 May
and 16 November from Uvalde. Although no collecting was done in
Uvalde during this study, Lu. diabolica were taken in resting and
biting collections at nearby Garner State Park (50 km north of Uvalde)
as early as 4 June and in light trap collections at the D'Hanis site
(50 km east of Uvalde) as early as 5 May and as late as 4 December.
Specimens of this species were taken at Garner State Park in human
biting collections by Young (1972) as early as May 17. Adults are
probably present in these locations throughout the frost free season,
with populations increasing over the summer months and reaching a peak
in the fall when they reach a nuisance level and are noticed by the
public. Parman (1919) apparently based appearance records on biting
activity. If so, this is consistent with the idea of a gradually
expanding population over the summer months. Surveys throughout the
year in and around Uvalde will be essential to confirm this idea.
Parman (1919) believed that Lu. diabolica never venture out of
hiding until well after sunset and never attack earlier than one hour
after sunset. Wilkerson (pers. comm., 1984) collected a female Lu.
diabolica biting him at 1400 hrs in full sunlight in mid July at
Canyon Lake, Comal County, Texas. During the 1982 research trip one
female was collected while biting the author's son at 1400 hrs in full
sunlight (temperature 380C, RH approximately 60%). These may be
isolated instances, but they indicate that perhaps the daily feeding
activity of Lu. diabolica is not as restricted as previously reported.
Lindquist (1936) reported the feeding period to be from 2000 hrs to
2400 hrs and Endris (1982) reported Lu. diabolica feeding during all
hours of darkness. Williams (1966b), in studies of biting rhythms of
ten anthropophilic sand flies in Belize, found the greatest period of
activity was between 0600 and 0659 hrs. After this small burst of
activity, the number of flies decreased. He said that the flies were
least active between 1400 and 1459 hrs (the hottest part of the day),
but that biting activity increased gradually from 1500 hrs onward.
Flies were not collected in appreciable numbers, however, until dusk
(1800-1859 hrs). He noted that numbers increased still further during
the early hours of darkness, reaching the peak of greatest activity
between 2100 and 2159 hrs. Thereafter, biting density diminished.
Consistent with these later reports, the peak feeding period of Lu.
diabolica in June 1982 was observed between one hour after sunset and
midnight (2130 to 2400 hrs). Killick-Kendrick and Rioux (1981)
described similar activity for Phlebotomus ariasi Tonnoir in the
Cevennes, France, and said it was probably triggered, at least
partially, by the rise in relative humidity as the temperature falls
at sunset. Parman (1919) said Lu. diabolica would not bite in total
darkness or in full moonlight. Although most biting collections in
June 1982 were taken in the presence of artificial light, sand flies
were also collected in lesser numbers while biting in total darkness,
having no apparent difficulty finding their host. They were also
observed biting outdoors in full moonlight.
Parman (1919) noted that the abundance of Lu. diabolica in the
fall was extremely variable, ranging from only one specimen attacking
in several nights, to 25 to 30 attacking each night for a short
period. This variability in numbers and peak time of attack was
observed both in the spring and fall during this study, and appears to
be strongly influenced by ambient temperatures, relative humidity and
air movement. Sand flies were observed in greatest numbers on hot
humid nights (270C or above and 70% or greater RH) with little or no
air movement. Not surprisingly, air movement seemed to affect their
presence the most, as numbers decreased dramatically with an increase
in wind velocity above 8 kph (5 mph). Foraging sorties or wave
attacks such as described previously did not occur except on hot,
humid and windless nights. Killick-Kendrick and Rioux (1981) reported
the peak biting activity of P. ariasi may be delayed because of wind
or suppressed by storms or a fall in temperature below about 160C.
Some workers have studied the effects of wind on the movement and
activity of sand flies but probably overestimated the wind speed at
which activity ceased (Killick-Kendrick and Rioux, 1981).
Blood-fed females have greater difficulty flying than unfed
females and apparently rest in protected sites for up to 24 hrs to
allow for diuresis and partial digestion of the blood-meal. This
accounts for the large number (31 to 46%) of blood-fed females taken
in latrine resting stations.
Both male and female Lu. diabolica are assumed to disperse
from their breeding sites, possibly in search of sugar or a blood-
meal. Their attraction to light draws them close to human dwellings
where other shorter range attractants, such as exhaled CO2
or body warmth, may aid them in finding a blood meal. This may also
be a genetically selected, triggered-sequence response, i.e., lights =
people = food. For whatever the reason, their response to light
contributes to the success of light-trapping, and several workers have
collected Lu. diabolica by this method (Young and Perkins, 1984).
Indeed the lighted latrines at Garner State Park and Fawcett Boy Scout
Camp functioned as giant light traps. The exposed positions of the
positive latrines was somehow important, since those at lower
elevations in more protected positions yielded nothing. Perhaps they
were more visible and attracted flies from considerable distances. It
may also be that the sand flies were seeking something other than a
blood meal, such as shelter. As was the case at the farm in D'Hanis,
CDC light traps may not be bright enough to attract sand flies and
must be augmented with CO2 (dry ice). The brighter lights of the
Shannon trap and the New Jersey light traps were apparently sufficient
to attract foraging flies.
The distance traveled by a sand fly to its host may depend upon
the habitat and the species. Estimates by other workers of distance
traveled to light ranges from 200 m in a neotropical forest (Chaniotis
et al., 1974), to as far as 2300 m in open habitat (Killick-Kendrick
and Rioux, 1981). These authors also suggested that the movement of
females is a hunting strategy and hence the difference in dispersal
distance between males and females. Since Lu. diabolica occurs in a
rather open habitat, it is likely that they disperse a considerable
distance (1000 m or more) from the unknown breeding site. Males
probably have a more limited flight range. The greater number of
males found in resting collections at Fawcett Boy Scout Camp may
have been an indication that the breeding site was close by. Mark-
release and recapture experiments will be necessary to determine the
actual extent of dispersal from the breeding site.
Lutzomyia anthophora. This species was first collected
while feeding on rabbits in Uvalde, Texas, in the type locality of Lu.
diabolica (Addis, 1945a). Easton (1968) collected them in Malaise
traps in Kinney and Presidio Counties, Texas. Young (1972) found Lu.
anthophora in the nest of the plains woodrat (Neotoma micropus) in San
Antonio, along the Rio Grande near Brownsville, Texas, and at Welder
Wildlife Refuge near Sinton, Texas. Endris (1982) also collected Lu.
anthophora from woodrat nests near Brownsville, Texas. Finding this
species at Garner State Park, Fawcett Boy Scout Camp, and D'Hanis
expands its known geographic distribution (Fig. 2-30). The close
association of Lu. anthophora with the woodrat, and the finding of
recently engorged females in the soft inner nest of the main Neotoma
den strongly suggest the preferred host to be the woodrat (Young,
1972). Endris (1982) fed females of this species on the following
anesthetized animals: woodrat, white footed mouse (Peromyscus
leucopus), Syrian hamster (Mesocricetus auretus), grey squirrel
(Sciurus carolinensis), white mouse (Mus musculus), guinea pig (Cavia
porcellus), domestic rabbit (Oryctolagus cuniculus), and opossum
(Didelphis marsupialus). Rodent reservoirs of leishmaniasis in the
Neotropics, but not in Texas, have been reported by several authors
(Bray, 1974a; Lainson and Shaw, 1979). Still, this possibility must
not be overlooked and should be investigated in future studies. Lu.
anthophora is not known to be anthropophilic, but Perkins (pers.
comm., 1984) reported that females in flourishing laboratory colonies
will occasionally bite humans.
Lutzomyia texana. Lutzomyia texana was described from
specimens collected in the nest of the leaf-cutting ant, Atta texana
(Buckley), in San Antonio, Texas. This species has been collected in
light traps at several sites throughout south central Texas (Fig.
2-31). Young (1972) reported that Lu. texana frequently inhabit
mammal burrows, especially those dug by armadillos, and that they were
collected from such places throughout most of the year. Efforts
to feed specimens on a variety of hosts in June 1982 and September
were unsuccessful. Young (pers. comm., 1984) believes that armadillos
may be the principal hosts. Lainson et al. (1979) isolated Leishmania
from armadillos in Brazil. There is no evidence of armadillo
reservoirs of cutaneous leishmaniasis in Texas; however, the
matter has never been investigated and should be given further
Lutzomyia vexator (Coquillet). Lutzomyia vexator is the most
widely distributed sand fly in the USA (Young and Perkins, 1984). It
was collected previously by Young (1972) at Garner State Park and
Fredricksburg, Texas, in light traps. These flies are reptile feeders
and are of no known medical or economic importance.
Lutzomyia new species. This new species, known from a single
female, raises to seven the number of sand fly species collected from
Texas. Future field surveys will be essential to collect more
specimens on which to base a valid description of the species and to
study its biology and host associations.
*= autochonous human case I
of cutaneous leishmaniasis!
0= known distribution of
Known geographical distribution of the sand fly
anthophora in Texas.
*= autochthonous human case
of cutaneous leishmaniasi
0= known distribution of
Known geographical distribution of the sand fly Lutzomyia
texana in Texas.
Based on the results of this field study and the records of other
workers, it can be said that Lu. diabolica is an abundant, widely
distributed sand fly species in south central Texas. As the only
known anthropophilic species known in the state, with peridomestic
habits, and having a distribution that roughly coincides with areas of
Leishmania endemicity, it is strongly implicated as the probable
vector of human cutaneous lesishmaniasis in Texas (Fig. 2-29).
Further evidence to support this is provided by experiments in which
laboratory-fed Lu. diabolica were shown capable of transmitting
L. mexicana from infected to uninfected hamsters by bite. These
experiments are discussed in detail in Chapter 4.
Endris et al. (1984) transmitted L. mexicana from infected to
uninfected hamsters with laboratory-bred Lu. anthophora. This led
them to suggest that L. mexicana could be maintained in wild rodent
populations by a non-anthropophilic species, such as Lu. anthophora,
and secondarily transmitted to man by a sympatric anthropophilic sand
fly, such as Lu. diabolica, thus implicating Lu. anthophora as a
possible accomplice in the transmission of human cutaneous
leishmaniasis in Texas. A further possibility that can not be ruled
out is that Lu. texana may likewise be implicated as an accomplice
vector when knowledge of its hosts and feeding behavior are revealed.
Processing and Maintenance of Wild-Caught Sand Flies
and Recovery of Eggs
The advantage of recovering eggs from wild-caught females in the
field is clear. They are much easier to handle than adults and
without significant mortality, allowing for field transportation of
F1 genetic strains for laboratory studies.
It was found that gravid females could be transported long
distances in oviposition vials in a vehicle when they were protected
from excessive heat and provided moisture periodically. One batch of
about 40 adult females was hand carried 1,000 miles on an airplane
with no mortality.
Oviposition records (Table 2-3) indicate that most of the females
did not deposit all their eggs and that the number of eggs laid per
female was quite variable (1-84 in June 1982; 15-76 in September
1983). Less variability in number of eggs deposited was observed in
1982. The higher percentage of females depositing ova during the fall
1983 trip as compared to spring 1982 (Table 2-3) is probably a
reflection of improved handling techniques and "laboratory"
facilities, and cooler temperatures. During June, 1982, the
"laboratory" was a screened camping shelter in which temperatures
remained about ambient. Mean temperatures at Garner State Park and
surrounding areas during that month were about 280C with average daily
maximum temperatures of about 350C. Relative humidity ranged between
about 50 and 95%. Special care such as adding ice to the polystyrene
cooler and addition of water to the plaster in the vials kept the
specimens alive until oviposition. In September, 1983, ambient
temperatures were lower (mean, 250C; maximum, 320C; RH about 65%) and
the "laboratory" facility consisted of an air-conditioned room that
was maintained at approximately 240C.
The mean gross egg production (eggs laid plus eggs retained) of
captured females was about the same for both trips. The high
percentage of females retaining eggs may simply reflect the stress of
captivity. The number of females retaining eggs in the laboratory
colony (after 13 generations) was much less and presumably
approximates what would be found in nature. Females that neither
deposited nor retained ova amounted to 3% of the catch in 1982 and 10%
in 1983. Whether this is evidence of possible gonotrophic
disassociation is only conjecture. Occurrence of the phenomenon to a
higher degree in the fall of the year is consistent with what has been
observed in other insects.
Differences in preoviposition interval, postcapture longevity, and
postoviposition longevity were also probably due to improved handling
techniques in the second year when it was cooler (Table 2-4).
Comparison of preoviposition longevity of females captured with and
without blood meals (Table 2-4) reveals no significant differences,
suggesting that females captured with blood meals had recently fed and
that blood-fed females do not linger at the latrine resting stations
more than a few hours.
At least 57% of the egg batches laid by wild-caught, blood-fed
females hatched. Since these females were not placed with males after
capture, it is obvious that at least this percentage of females were
inseminated prior to capture. Sperm were observed in the spermathecae
of many of the dissected females, but spermathecae were not always
visible in dissections.
Accessory Glands and Parity
According to Chapman (1971), accessory glands in most insects
arise from the genital chamber or vagina. Their function varies in
different insects, but they commonly produce a substance for attaching
the eggs to the substrate. This material may also serve as a
protective covering for the eggs. There is considerable confusion,
even controversy, in the literature over the value of accessory gland
examination as an indicator for determine parity in sand fly
populations. Adler and Theodor (1935) stated that it was possible to
distinguish between blood-fed and unfed females of the palearctic sand
fly, P. perniciosis Newstead, by examining the accessory glands since
granules never appeared in the accessory glands unless the female had
had a blood meal. A few days after a small blood meal they were full
of granules. They further stated that during egg laying most of the
granules were passed out with the eggs, but some granules remained and
could be seen in dissections. The presence of granules in the
accessory glands of female P. perniciosus with an empty alimentary
tract was the only morphological feature that distinguished parous
females from newly hatched ones. They also stated that the granules
were formed independent of copulation, depending only on a blood-meal
in P. perniciosus.
Adler and Theodor (1957) referred to the indicative value of the
accessory glands for sand flies in general. Garnham and Lewis (1959)
noted high proportions of dissected sand flies with granules in the
accessory glands and concluded that some of the flies might be
nullipars secreting granules. They suggested studying the value of
these organs as indicators of nulliparous flies in Belize. Lewis and
Minter (1960) examined ovaries and accessory glands of some tropical
African sand flies, in Kenya, and found that when the ovaries were
small, residual secretions in the accessory glands were useful in
recognizing most parous females. Johnson and Hertig (1961) found that
some Panamanian sand flies secreted accessory gland granules before
taking blood and discharged them at various times afterwards. Adler
and Mayrink (1961) observed dark brown fluid in the accessory glands
of blood-fed Lu. longipalpis (Lutz and Neiva) and noted changes in
glands of five other species. Johnson et al. (1963) found that in
laboratory-reared Panamanian sand flies, granules may be present or
absent without regard to parous or nulliparous condition of the
female. They concluded that examination of the accessory glands does
not aid in determining whether females of Panamanian sand flies are
parous or nulliparous. Minter (1964) believed that the accessory
glands were useful in estimating parous rates in Kenya. Lewis (1965)
examined accessory glands of five species of Lutzomyia in Belize and
concluded that the glands were unreliable for indicating whether or
not a fly was parous. Chaniotis and Anderson (1968) studied
laboratory-bred females of three California species and found no
granules in accessory glands of 150 nullipars. Scorza et al. (1968),
on the other hand, concluded that accessory gland granules were
unreliable for recognizing parous females in Venezuela. Lewis et al.
(1970) reconsidered a statement made earlier by Lewis (1965) regarding
the indicative value of accessory glands by stating that the glands
are quite useful for recognizing parous females of many Old World and
some New World species. They reported that errors in interpretation
are caused by irregular secretions, loss of secretion, the effect of
blood-meals, and the parasite Monocystis (Ascocystis). Ward (1974)
found that 82.05% of laboratory-reared Lu. longipalpis developed
granules, but no eggs when kept for seven days without blood or sugar.
He questioned the value of accessory glands in determining parity in