|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
Aedes albopictus OVIPOSITION AND LARVAL DENSITY, DEVELOPMENT, AND
INTERACTIONS WITH Wyeomyia spp. IN EXOTIC BROMELIADS OF SOUTHERN
ROBYN R. RABAN
A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE
UNIVERSITY OF FLORIDA
Robyn R. Raban
To my parents, Bill and Judy Raban and my grandfather, Bill Phillips for their support for all my
I would like to thank L.P. Lounibos for his advice and support. I thank G. F. O'Meara for
his assistance in locating field sites, and his enthusiastic support for my research. I thank J.H.
Frank for helpful suggestions on my research; N. Nishimura for help with my technical
problems; and R. Escher for his mosquito rearing guidance and for his provision of eggs for my
research. Finally, I would like to thank the faculty, staff, and graduate students at FMEL, most
of whom at one time or another, provided invaluable help to my research.
TABLE OF CONTENTS
ACKNOWLEDGMENT S .............. ...............4.....
LI ST OF T ABLE S ................. ...............7................
LI ST OF FIGURE S .............. ...............9.....
AB S TRAC T ........._. ............ ..............._ 1 1...
1 Aedes albopictus AND Wyeomyia spp. IN THE EXOTIC BROMELIADS OF
SOUTHERN FLORIDA. .............. ...............13___ .......
Aedes albopictus in the United States............... ...... ........... ...........1
Wyeomyia mitchellii and Wyeomyia vanduzeei in the United State s ................ .................1 4
Bromeliaceae as a Habitat ................ ......... .. ........... .......1
Interactions of A. albopictus and Wyeomyia spp. in Bromeliad Phytotelmata ................... ....1 7
2 INTERFERENCE COMPETITION AS A POTENTIAL MECHANISM FOR THE
REDUCTION OF GROWTH AND SURVIVORSHIP OF Aedes albopictus IN THE
PRESENCE OF Wyeomyia spp. IN THE BROMELIADS OF SOUTHERN FLORIDA. ....20
Introducti on .................. ...............20._ ___......
M materials and M ethods .............. ...............24....
Chemical interference................. .. .. .. ..............2
Surface Area to Volume Ratio Encounter Competition Experiment .............. ................26
Habitat Complexity Experiment.................. .... ................2
Resource Dependent Location of A. albopictus and Wyeomyia spp. ............. ................28
Re sults................... ..... ...............30.......
Chemical Interference ................ ....... .............3
Surface Area to Volume Ratio Experiment ................. ........._. ...... 30__. ...
Habitat Complexity Experiment ......................... ......... .... ........
Resource-Dependent Foraging Location ...._. ................. ........___.........3
Discussion ................. ...............32....... ......
3 Wyeomyia spp. AS POTENTIAL PREDATORS OF Aedes albopictus IN
BROMELIADS IN SOUTHERN FLORIDA .............. ...............53....
Introducti on ........._ ....... __. ...............53....
M materials and M ethods ................... ........... .. .. ..............5
Testing for Density Dependent Predation by Wyeomyia spp. ............. .....................5
Testing for Resource Dependent Predation by Wyeomyia spp............... ..................5
Predation Observation .............. ...............55....
Re sults ................ ...............55.................
Discussion ................. ...............56.................
4 DO Wyeomyia spp.. LARVAE INHIBIT EGG HATCH OF Aedes albopictus? ..................59
Introducti on ................. ...............59.................
M materials and M ethods .............. ...............60....
Re sults ................ ...............61.................
Discussion ................. ...............62.................
5 FIELD STUDIES ON Aedes albopictus AND Wyeomyia spp. IN EXOTIC
BROMELIADS OF SOUTHERN FLORIDA. ......___. ...... .___ .... _. ...........6
Introducti on ........._ ....... __. ...............68....
M materials and M ethods .............. .. .. ..... ..... ... ...............6
Differences in the Density ofA. albopictus and Wyeomyia spp. within N.
spectabilis and B. pyramnidalis ............_. ...._.. ...._... ...........6
Canopy Effects on Density ............._. ...._... ...............71...
Re sults............._. ...._... ...............72....
Discussion ............._. ...._... ...............73....
6 OVIPOSITION AND LARVAL DEVELOPMENT OF Aedes albopictus IN TWO
EXOTIC SPECIES OF BROMELIAD IN SOUTHERN FLORIDA .................. ...............91
Introducti on ................. ...............91.................
M materials and M ethods .............. ...............94....
Oviposition Experiment. ............. ... .... ...............94..
Larval Development and Survivor ship Experiment ............. ..... ...............96
Re sults............. ..... .. ...............96..
Oviposition Experiment. ............. ... .... ...............96..
Larval Development and Survivor ship Experiment ............. ..... ...............96
Discussion ............. ..... ...............97...
7 CONCLUSIONS .............. ...............100....
LI ST OF REFERENCE S ............. ..... ._ ...............103..
BIOGRAPHICAL SKETCH ............. .....__ ...............116..
LIST OF TABLES
2-1 Means and standard errors for the concentrations of ammonia and pH pre- and post-
experiment. ........._.._.. ...._... ...............36....
2-2 Analysis of variance for A. albopictus average instar based on the presence or
absence of Wyeomyia spp. and the surface:volume ratio of the artificial plant and the
interaction of these two variables. ........._ ....__ ......__ ......_ ............37
2-3 Analysis of variance for A. albopictus average survival based on the presence or
absence of Wyeomyia spp., and the surface area:volume ratio of the plant (size) and
the interaction of these two variables. ..........._... ...............38.....__.._....
2-4 Multivariate analysis s of variance table for feeding location ................. ............. .......3 9
2-5 Analysis of variance for the effects of food level presence or absence of Wyeomyia
spp. on the location of A. albopictus ................. ...............40......__. ..
2-6 Analysis of variance for the effects of food level and presence or absence ofA.
albopictus on the location of Wyeomyia spp............... ...............41..
3-1 Number of Wyeomyia examined in predation experiments. ............... ...................5
5-1 Analysis of variance for the density of mosquitoes per plant based on month (month
of collection), site, and bromeliad species ................. ........._. ......77.........
5-2 Analysis of variance for the density of mosquitoes per mL based on month (month of
collection), site, and bromeliad species .............. ...............78....
5-3 Analysis of variance for the density of each mosquito species per mL of water in
each bromeliad species. ............. ...............79.....
5-4 Mean densities of mosquitoes (mos/plant) in B.pyramidalis and N. spectabilis based
on location within the plant and plant species. ............. ...............80.....
5-5 Analysis of variance table for the effects of month and site on mos/mL and
mos/pl ant. ........._.._.. ...._... ...............8 1....
5-6 Analysis of variance results for the effect of month and sites on mosquito
abundances (mos/plant) by bromeliad species ................. ...............82........... ...
5-7 Average (+SE) amount of water extracted from each plant species. ........._.._... ........._....83
5-8 A nested ANOVA table for effects of Oak vs. Palm canopy and bromeliad species
(B. pyramnidalis vs. N. spectabilis) on Wyeomyia spp. densities by mos/plant. ........._......84
5-9 Means and SE of densities ofA. albopictus and Wyeomyia spp. by location within
plants under oak and palm canopies. ............. ...............85.....
LIST OF FIGURES
2-1 Construction of the squares for the habitat complexity experiment. ............. .... ........._...42
2-2 The folded shape of the high complexity and low complexity treatments and the
orientation of the squares within the cones ................. ...............43........... ..
2-3 Average instars ofA. albopictus after 48 h exposure to Wyeomyia spp. through a
dialysis membrane and without Wyeomyia in the control. ............. .....................4
2-4 Mean survivorship of A. albopictus after 48 h exposure to Wyeomyia spp. through a
dialysis membrane and without Wyeomyia spp. in the control. The error bars are
standard error. ............. ...............45.....
2-5 The average instar & SE ofA. albopictus in the absence or presence of Wyeomyia
spp. in four plant sizes.. ............ ...............46.....
2-6 The average survivorship a SE of A. albopictus in the absence or presence of
Wyeomyia spp. in four plant sizes. ................. ...............47....____....
2-7 The average instar (+SE) of A. albopictus in two levels of habitat complexity and
three levels of surface area to volume ratio. ............. ...............48.....
2-8 Location ofA. albopictus in the presence and absence of Wyeomyia .............. ................49
2-9 Location of Wyeomyia spp. in the presence and absence ofA. albopictus. .......................50
2-10 Location ofA. albopictus by food concentration and the presence and absence of
Wyeomyia spp... ........... ...............51......
2.11 Location of Wyeomyia spp. by food concentration and the presence and absence ofA.
albopictus. ................. ...............52....... ......
4-1 Proportion hatch of viable eggs & SE in each treatment..........._.._.. ......._.._........._..66
4-2 Proportion eggs damaged +SE per treatment. ........._..._.._ ...._._ ....._._. .........6
5-1 M ap of collection sites. .............. ...............86....
5-2 Location of water samples. ........._..._._ ...............87.._.._._ ....
5-3 The monthly mean densities of Wyeomyia spp. and A. albopictus in N. spectabilis
and B. pyramnidalis at each site. ............. ...............88.....
5-4 The monthly mean densities of Wyeomyia spp. and A. albopictus in N. spectabilis at
each site............... ...............89..
5-5 The monthly mean densities of Wyeomyia spp. and A. albopictus in B. pyramnidalis at
each site............... ...............90..
Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Science
Aedes albopictus OVIPOSITION AND LARVAL DENSITY, DEVELOPMENT, AND
INTERACTIONS WITH Wyeomyia spp. INT EXOTIC BROMELIADS OF SOUTHERN
By Robyn R. Raban
Chair: L. P. Lounibos
Maj or Department: Entomology and Nematology
Wyeomyia mitchellii and Wyeomyia vanduzeei, are indigenous mosquitoes that inhabit
exotic and native bromeliads in southern Florida. In the mid-1980s Aedes albopictus invaded
Florida where its immature stages occupied artificial and natural containers, including
bromeliads. Previous studies have shown reduced abundance of A. albopictus in exotic
bromeliads co-occupied by Wyeomyia spp. immatures, and reduced larval growth and
survivorship in the presence of late instar Wyeomyia spp.
Our study examined chemical interference, encounter competition, predation, and egg
hatch inhibition as potential interactions between Wyeomyia spp. and A. albopictus in exotic
bromeliads. Exposure to waste products of Wyeomyia spp. did not affect the growth or
survivorship ofA. albopictus. Experimental alterations of surface area to volume ratios and
habitat complexity did not affect the decreased growth and survivorship ofA. albopictus seen in
the presence of Wyeomyia spp. Behavior experiments showed that A. albopictus larvae change
their location within a container in the presence of Wyeomyia spp. This change in location may
indicate a response ofA. albopictus to encounters with Wyeomyia spp. Wyeomyia spp. showed
no evidence of predation on A. albopictus, but fourth instar Wyeomyia spp. were shown to inhibit
the egg hatch of A. albopictus.
Field surveys and lab experiments were conducted on A. albopictus and Wyeomyia spp.
in Neoregelia spectabilis and Billbergia pyramida~lis bromeliads. Higher densities of Wyeomyia
spp. were found in B. pyramnidalis, while higher densities ofA. albopictus were found in N.
spectabilis. Correlations between the number of mosquitoes per plant and the volume of water
extracted were significant for Wyeomyia spp. and A. albopictus. Overhead tree canopy type
influenced Wyeomyia spp. larval densities in bromeliads. A. albopictus females oviposited
preferentially in N. spectabilis, and A. albopictus developed faster when exposed to Wyeomyia
spp. within N. spectabilis as compared to the same exposure in B. pyramnidalis.
Encounter competition with Wyeomyia spp. larvae is the most probable mechanism
reducing the growth and survivorship of A. albopictus in exotic bromeliads in southern Florida,
although other interactions may also influence the relative abundances of these mosquito species.
Aedes albopictus AND Wyeomyia spp. INT THE EXOTIC BROMELIADS OF SOUTHERN
With expanding global transport and commerce, humans have increased the distributions
of many plant and animal species, often to novel locations. While some of these species fail to
establish in these new locations (Mack et al. 2000), some become successful biological invaders
Aedes albopictus in the United States
Since its establishment in North America in the 1980s (Sprenger and Wuithiranygool
1986), Aedes albopictus has invaded various habitats that have previously been occupied by
other mosquito species. It has expanded its range throughout the southeastern and midwestern
United States by its transport in used tires (Hawley 1988). In Florida, A. albopictus was first
discovered in 1984, and is currently found throughout most of the state (O'Meara et al. 1995a,
Juliano et al. 2004). Invasions by A. albopictus have been much discussed (Lounibos 2002), and
this mosquito species continues to colonize new areas of western (Aranda et al. 2006) and
eastern Europe (Klobucar et al. 2006).
Aedes albopictus is one of the commonest mosquitoes in natural and artificial containers in
the eastern United States and Brazil (Lounibos 2002). Aedes albopictus larvae are often found
inhabiting man-made containers, but can also be found in natural phytotelmata such as treeholes
(Novak et al. 1993) and bromeliads (Lounibos et al. 2003, O'Meara et al. 2003). Displacement
by A. albopictus of mosquitoes that traditionally occurred in these habitats, such as Aedes aegypti
(Juliano 1998), and Culex pipiens (Carrieri et al. 2003) has been seen in some cases.
To date, considerations of the interactions of A. albopictus with its competitors have
included mating interference (Nasci et al. 1989), egg hatch inhibition (Edgerly et al. 1993,
Edgerly et al. 1991), larval resource competition (Juliano 1998, Braks et al. 2004, Griswold and
Lounibos 2005), predator-mediated competition (Gubler 1971, Lounibos et al. 2001, Griswold
and Lounibos 2005), and intraguild predation (Edgerly et al. 1999). Yet, so far competitive
exclusion ofA. aegypti by A. albopictus has not occurred mainly due to suspected differences in
the egg desiccation tolerances (Costanzo et al. 2005b), or differences in macrohabitat preferences
(Lounibos et al. 2001, Costanzo et al. 2005a). In some cases though, local extinction of A.
aegypti occurred after the arrival ofA. albopictus (O'Meara et al. 1993).
Aedes albopictus have been infected with over 22 species of arboviruses (Moore and
Mitchell 1997). In North America, A. albopictus have been found infected in nature with
LaCrosse (Gerhardt et al. 2001), Eastern Equine Encephalitis (EEE) (Mitchell et al. 1992), and
dengue viruses (Ibafiez Bernal et al. 1997). Therefore, there are potential health concerns related
to A. albopictus populations.
Wyeomyia mitchellii and Wyeomyia vanduzeei in the United States
Wyeomyia mitchellii ranges throughout Mexico (Diaz-Naj era and Vargas 1973), the
Caribbean region (Belkin and Heinemann 1975, Shroyer 1981), central and southern Florida
(Darsie and Ward 2005), and Hawaii (Shroyer 1981). This bromeliad specialist mosquito is
commonly found co-occurring in Florida with another congeneric bromeliad specialist,
Wyeomyia vanduzeei Dyar and Knab. Wyeomyia vanduzeei ranges through central and southern
Florida (Darsie and Ward 2005), and much of the Caribbean region (Belkin and Heinemann
Before the introduction ofA. albopictus, W. mitchellii and W. vanduzeei were the most
common day-biting mosquito species in the mixed oak forests of southern Florida (Edman and
Haeger 1977). Immatures of these Wyeomyia spp. are historically the most common mosquitoes
in the epiphytic bromeliad phytotelmata of southern Florida (Fish 1976, Frank 1983, Frank and
O' Meara 1985).
Wyeomyia vanduzeei and W. mitchellii are diurnal species with peak diel biting activity
occurring just prior to sunset (Edman and Haeger 1977) and oviposition occurring in the late
daylight hours (Frank et al. 1985). In a study conducted by Edman and Haegar (1977), both W.
vanduzeei and W. mitchellii adults appeared to be generalist feeders, feeding on species of
rabbits, deer, and birds. Wyeomyia vanduzeei are facultatively autogenous, while W. mitchelli
are not (O'Meara 1979), and their larvae may be slightly segregated as W. vanduzeei lay more
eggs in sunny versus shady locations, and these two mosquito species show preferences for
different plants (Frank and O'Meara 1985).
Although Wyeomyia spp. have not been found to vector arboviruses in Florida, Venezuelan
Equine Encephalitis virus has been found in W. mitchellii (Scherer et al. 1971) and Ilheus and
Wyeomyia viruses have been found in W. vanduzeei (Srihongse and Johnson 1965, 1967)
Bromeliaceae as a Habitat
Phytotelmata are water bodies that are held in plant structures such as flowers, leaf axils, or
tree holes (Maguire 1971, Fish 1983, Clements 1999). The family Bromeliaceae contains many
species which impound water in their leaf axils and are referred to as tank bromeliads (Frank
1983). There are over 2000 bromeliad species in the Americas (Frank 1983), with 16 species of
bromeliads native to Florida, 7 of these being tank bromeliads that hold water throughout most of
the year (Fish 1976).
While bromeliads obtain energy from photosynthesis, the nutrients they sequester for
photosynthesis are obtained in multiple ways. Tank bromeliads can absorb nutrients through
scales on the shoot called trichomes (Pittendrigh 1948, Benzing et al. 1976, Benzing 2000), or by
way of interfoliar roots (Pittendrigh 1948). Frank (1983) classifies nutrient acquisition by tank
bromeliads into two categories, dendrophilous nutrition, which is rain throughfall from trees
often collected by epiphytic bromeliads, and anemophilous nutrition, which are wind blown
nutrients. Frank (1983) and Benzing (2000) also describe the release of nutrients provided by the
wastes of faunal tank inhabitants which may contribute to the nutrients available to the plant.
In bromeliads, mosquitoes may encounter a wide variety of other unrelated species. To
date, bacteria, fungi, algae, protozoa, platyhelminthes, rotifers, gastrotrichs, ostracods,
nematodes, oligochaetes, crustacea, acarids, Hemiptera, Coleoptera, Culicidae, Psychodidae,
Syrphidae, Ceratopogonidae, Chironomidae, Sarcophaghidae, Tabanidae, Periscelidae,
Tipulidae, Muscidae, Sciaridae, and odonates have all been identified within bromeliad
phytotelamata (Maguire 1971, Fish 1976, Frank 1983). Some chironomids, muscids,
periscelids, and turbellarians are predatory and feed upon smaller organisms and mosquito eggs,
while some psychodids and sciarids feed on the submerged leaf litter in the bromeliad (Fish
1976). Bromeliads vary in their faunal compositions by the species of bromeliad and
environmental conditions, such as sunlight and humidity (Laessle 1961, Fish 1976).
Since the popularization of exotic bromeliads as ornamental plants (Edman and Haeger
1977), there has been an increase in exotic bromeliad habitat for Wyeomyia spp. in Florida. Yet,
with the arrival of the Mexican bromeliad weevil, M\~etamasiust~~~ttt~~ttt~~ callizona, there has been a decline
of the native bromeliad Tillandsia~~ll1~~~~111~~~ utriculata (Frank and Thomas 1994), the common native
phytotelmata of W. mitchellii and W. vanduzeei (Fish 1976). Exotic bromeliads, which have
become another common habitat of these two species, hold more water than the native bromeliad
species and have a large central water-holding tank, uncommon in native bromeliads (O'Meara
et al. 2003).
The two bromeliads which were the focus of most of this study are Billbergia pyramnidalis,
the summer torch bromeliad, and Neoregelia spectabilis, the painted fingernail plant. Billbergia
pyramnidalis and Neoregelia spectabilis are native to Brazil (Frank et al. 1988), and can both live
as either an epiphyte or as a ground dwelling species. Neoregelia spectabilis is the larger of the
two species and tends to be a dark green to dark purple color with bright purple tips. The
inflorescence of N. spectabilis is nidulate, meaning it is nested into the central tank of the plant.
Billbergia pyranzida~lis is a light green color and has a stalked inflorescence.
Interactions of A. albopictus and Wyeomyia spp. in Bromeliad Phytotelmata
There is evidence that Wyeonzyia spp. larvae exhibit negative effects on A. albopictus
within bromeliads. Possibly due to the freezing temperatures, or the absence of native
bromeliads in northern Florida, W. nitchelliii and W. vanduzeei are infrequently found north of
Orlando (O'Meara et al. 2003). In the absence of Wyeonzyia spp. in northern Florida, A.
albopictus has become a common mosquito in bromeliads, but in southern Florida where
Wyeonyia spp. are abundant, A. albopictus individuals are scarce in bromeliads (O'Meara et al.
1995b, Lounibos et al. 2003). A study conducted in Hawaii determined that in the absence of
Wyeonzyia spp., A. albopictus dominated leaf axil phytotelmata (Shroyer 1981).
This study attempts to further investigate the results of Lounibos et al. (2003). In that
study, 1st instar A. albopictus were exposed to varied quantities of leaf litter with 1st or 4th instar
Wyeonzyia spp. in B. pyra~nidalis bromeliads. Aedes albopictus experienced a reduction in
survivorship with reduced leaf litter and the presence of 4th instar Wyeonzyia spp., but the
presence of Wyeonzyia spp. accounted for most of the reduction in survivorship. The mean
growth stage of A. albopictus was only affected by the presence of 4th instar Wyeonzyia spp. with
no effect due to leaf litter. 1st instar Wyeonzyia spp. had no effect on the growth or survivorship
of A. albopictus.
Lounibos et al. (2003) hypothesized that the reduction in growth and survivorship of A.
albopictus in the presence of Wyeonzyia spp. was attributable to interference competition.
Lounibos et al. (2003) suggested that because the addition of leaf litter did not reduce the
negative effects of Wyeomyia spp. on A. albopictus, this interaction is independent of resources.
Interference competition is categorized into overgrowth, chemical, territorial, and encounter
(Schoener 1983). Overgrowth interference competition occurs only in sessile organisms, and
mosquito larvae are not known to exhibit territorial interference behaviors, so these mechanisms
are not considered in the possible competitive interactions between A. albopictus and Wyeomyia
Chemical interference in mosquitoes results from the buildup of larval waste products that
lead to an environmental deterioration which can effect the growth and survivorship of
individuals (Bedhomme et al. 2005). Aedes albopictus has been found to be subj ect to chemical
interference competition with the mosquito Tripteroides bamnbusa (Sunahara and Mogi 2002).
Chemical interference has also been demonstrated in Aedes sierrensis (Broadie and Bradshaw
1991), A. aegypti (Dye 1984), and Culex sitiens (Roberts 1998).
Encounter competition usually occurs when physical contact between organisms causes a
reduction in the feeding efficiency of individuals (Anholt 1990). Encounter competition is
known among A. sierrensis (Broadie and Bradshaw 1991), A. aegypti (Dye 1984) and Culex
sitiens (Roberts 1998). Other mosquitoes such as Ochlerotatus cantans (Renshaw et al. 1993)
are hypothesized to also exhibit encounter competition effects.
Intraguild predation is another important regulatory mechanism in mosquito populations.
Some mosquito species are facultative predators or cannibalistic on smaller instars (Hinman
1934, Reisen and Emory 1976, Koenekoop and Livdahl 1986, Koenraadt et al. 2003, Edgerly et
al. 1999, Koenraadt et al. 2004). Egg hatch inhibition occurs when mosquito larvae delay hatch
of Aedes eggs. Egg hatch inhibition could increase the time to reproduction, which may reduce
population growth and increase mortality in the dormant egg stage of some species (Livdahl and
Edgerly 1987). Some Aedes species exhibit varying degrees of egg hatch inhibition (Livdahl et
al. 1984, Koenekoop and Livdahl 1986, Livdahl and Edgerly 1987).
Thus, the purpose of the following experiments on A. albopictus and Wyeomyia spp. is to
determine the mechanism of reduction of growth and survivorship of 1st instar A. albopictus in
the presence of 4th instar Wyeomyia spp. in exotic bromeliads. As mentioned above, chemical
interference competition, encounter competition, predation, and egg hatch inhibition are potential
competitive mechanisms in mosquito populations. Previously none of these mechanisms have
been studied between Wyeomyia spp. and A. albopictus within bromeliads, and thus these
interactions are the focus of a set of experiments in this study.
Furthermore, in attempt to add to the current understanding of the occupancy by A.
albopictus and Wyeomyia spp. of bromeliads in southern Florida (O'Meara et al. 1995b,
Lounibos et al. 2003) multiple collections were made from two common ornamental bromeliad
species B. pyramnidalis and N. spectabilis. Subsequent oviposition and larval competition
experiments were conducted with these two bromeliad species to explain the patterns of
mosquito abundance observed from field collections.
INTERFERENCE COMPETITION AS A POTENTIAL MECHANISM FOR THE
REDUCTION OF GROWTH AND SURVIVORSHIP OF Aedes albopictus IN THE
PRESENCE OF Wyeomyia spp. IN THE BROMELIADS OF SOUTHERN FLORIDA.
Competitive interactions can be a maj or factor in determining community structure.
Exploitative and interference competition are the two main mechanisms for interactions between
individuals within a population (Schoener 1983). Exploitative competition has been the focus of
many ecological studies, but interference competition often has a greater influence over species
distribution and abundance (Case and Gilpin 1974). Schoener (1983) subcategorized
exploitative competition into consumptive competition, which involves food resource
competition, and preemptive competition for resource space. The subcategories of interference
competition include (i) overgrowth, which involves individuals growing over one another and
thus depriving their competitors of resources; (ii) chemical, which deals with toxins produced by
some individuals harming other individuals; (iii) territorial, in which individuals aggressively
defend units of space to the detriment of the competitors, and (iv) encounter competition in
which interactions between mobile competitors cause harm by fighting, predation, or physical
interference. Overlap is common in exploitative and interference competition (Schoener 1983).
For example, interference competition affects competitors by changing the rate of resource
exploitation, or in some cases resulting in injury or death, which in turn affects population
growth (Case and Gilpin 1974).
Many plants, such as grasses (Javaid et al. 2005, Singh et al. 2005) and pine trees
(Nektarios et al. 2005), release allelopathic chemicals that create an unfavorable environment for
possible plant competitors. In some cases, the allelopathic chemicals released by plants are the
main determinants of the plant community structure (Rasmussen and Rice 1971), and are
therefore a maj or competitive influence in the community. Bacteria also release allelopathic
chemicals to limit the growth of their neighbors (Riley and Gordon 1999).
Allelopathy has not been identified in animals, but animals experience chemical
interference caused by waste materials from other individuals. These waste products, such as
ammonia and nitrates, cause a deterioration of the environment leading to a reduction in the
growth and survivorship of other individuals (Bedhomme et al. 2005). For instance, intraspecific
chemical interference competition has been shown among sea lamprey larvae (Rodriguez-Mufioz
et al. 2003), Parantecium (Gill 1972), tadpoles (Griffiths et al. 1991) and larvae of the
mosquitoes A. aegypti (Bedhomme et al. 2005, Dye 1984) and Culex pipiens nzolestus (Ikeshoji
et al. 1976).
Density dependent chemical inhibition of larval growth has been demonstrated in other
species of Aedes mosquitoes. Broadie and Bradshaw (1991) determined from a laboratory
experiment that intraspecific chemical interference competition influenced pupation success,
pupal weight and development time in Aedes sierrensis. Chemical interference has also been
shown to affect the growth of A. albopictus. Sunahara and Mogi (2002) determined that A.
albopictus experienced a reduction in survivorship and pupation success from interspecific
chemical interference competition with Tripteroides ba~nbusa in bamboo stumps.
Chemical inhibition of larval growth may be less common in other genera of mosquitoes.
Wyeonzyia snzithii exhibit density dependent increases in development time and reductions in
survivorship and pupal weight, but neither encounter competition, chemical competition, nor
cannibalism were found to contribute to these density dependent changes in fitness (Broberg and
While chemical inhibition of growth does occur within larval communities, encounter
competition is often found to be the predominant interference mechanism. Dye (1984) found that
some strains ofA. aegypti demonstrated intraspecific chemical interference, but chemical effects
were small compared to those of encounter competition. Roberts (1998) also found that chemical
inhibition of growth occurred among Culex sitiens conspecifics, but that encounter competition
was the stronger of the two competitive factors.
Encounter competition was defined by Brian (1956) as the ability of an organism to harm
another organism by directly attacking it, or indirectly by damaging its food resources or
blocking its access to those resources. Direct attack encounter competition has been found to
occur in mosquito larvae and often alters their survivorship. Among some fi1ter-feeding
mosquito species, 1st instar larvae may be attacked and often killed by fourth instar conspecifies
in laboratory microcosms (e.g., Reisen and Emory 1976, Koenekoop and Livdahl 1986,
Koenraadt et al. 2003). Mosquitoes may experience encounter competition when either high
densities of individuals cause frequent encounters with one another (Broadie and Bradshaw
1991, Roberts 1998) or larvae frequently encounter larger individuals of later instars (Dye 1984,
Broadie and Bradshaw 1991). The combination of high densities and larger individuals can have
the biggest effect on early instar mosquitoes (Broadie and Bradshaw 1991).
In many cases, physical contact between organisms can lead to a reduction in feeding
efficiency or other metabolic costs (Anholt 1990). Physical contact was implicated as the cause
of reduced feeding rates of A. sierrensis (Broadie and Bradshaw 1991) and Anopheles gamnbiae
(Koenraadt et al. 2003) at high densities.
Reductions in feeding efficiency can cause alterations in life history characteristics by
increasing larval mortality and development time and decreasing adult size. Aedes cantans
experienced a reduction in adult size and an increase in larval mortality, which were
hypothesized to be due to contact competition (Renshaw et al. 1993).
Previous studies have tried to vary the effects of encounter competition by altering the
water volume (Dye 1984), the surface to volume ratio (Broadie and Bradshaw 1991, Roberts
1998), or habitat complexity (Broadie and Bradshaw 1991). While experiments with alterations
in the surface area to volume ratio did not significantly alleviate the effects of competition by
changing the encounter rate between individuals (Broadie and Bradshaw 1991, Roberts 1998),
there were successes with the alteration of water volume (Dye 1984) and habitat complexity
(Broadie and Bradshaw 1991).
Habitat complexity affects the levels of competition and predation among species (Hixon
and Menge 1991, Hixon and Jones 2005). With increasing structural complexity of a habitat, the
number of competitive refuges increases (MacArthur and Levins 1967, Finke and Denno 2002),
and the number of physical encounters between predator and prey decreases (Murdoch and
Oaten 1975). With more physical encounters in habitats of decreased complexity, the occurrence
of intraguild predation increases (Marshall and Rypstra 1999, Roda et al. 2000, Finke and Denno
2002). In fact, most studies conducted on the effects of habitat complexity on competitive
interactions have involved predator-prey interactions (e.g., Crowder and Cooper 1982, Schneider
1984, Diehl 1992, Babbitt and Tanner 1998, Alto et al. 2005), although some studies have found
benefits to competitors as well as prey populations with increasing habitat complexity (Almany
2004). As with predator-prey interactions, an increase in habitat complexity should decrease
encounters between potential competitors. Therefore, structural complexity may provide refugia
from physical encounters with non-predatory species.
Behavioral changes in organisms that experience encounter competition are common.
Anholt (1990) observed noticeable behavioral changes in feeding behavior of damselflies when
exposed to different interspecific encounter rates. As mentioned before, physical encounters
with other individuals can cause changes in feeding efficiency, and thus there should be a change
in feeding behavior of individuals influenced by encounter competition.
Lounibos et al. (2003) found that A. albopictus had a reduction in growth and survivorship
when exposed as 1st instars to 4th instars of Wyeomyia spp. in bromeliads and hypothesized
interference competition as the mechanism for this effect. In this chapter, chemical interference
competition and encounter competition are examined as potential types of interference to explain
the negative effect of Wyeomyia spp. on larvae ofA. albopictus.
In this chapter, one experiment was conducted to determine whether the waste products of
Wyeomyia spp. reduce the growth or survivorship of A. albopictus. Two experiments were
conducted to determine whether physical contact with the larger 4th instar Wyeomyia spp.
affected the growth and survivorship ofA. albopictus. The first of these experiments varied the
surface area to volume ratio of artificial bromeliads in an attempt to change the frequency of
encounters between the species. The second experiment varied habitat complexity as an
alternative method of changing the frequency of encounters between species. Finally, the last
experiment attempted to elucidate changes in feeding location ofA. albopictus in response to the
presence of Wyeomyia spp..
Materials and Methods
Twenty-five 1st instar colony-raised A. albopictus were added to 1000 Dalton, 31 mm
diameter Spectra/Por Biotech Cellulose Ester dialysis tubes that contained 25 mL of sieved
bromeliad water and 0. 1 g of chopped live oak leaves (Quercus virginiana). The leaves were
dried at 68 OC for 48 h before weighing. The bromeliad water was collected from approximately
60 plants, and sieved through a 130 Clm gauge mesh screen to remove detritus and macroscopic
organisms. All sieved water was mixed into one large bucket, and then covered with a plastic
bag for 2 days.
The 1000 Dalton dialysis tubing was chosen for this experiment because its pore size
allows waste products such as nitrates and ammonia to pass through, while excluding larger
compounds such as food particles. Each tube was cut to 11.5 cm in length, which created
approx. 8 X 3.1cm of aquatic habitat for the l't instar larvae, and excess tubing of approximately
3.5 X 3.1cm was left over for folding the ends and for suspending the tubing. A butterfly clip
strung with wire was attached to the tubing to suspend it in the center axil of a Billbergia
pyramnidalis bromeliad. The butterfly clip also kept the tubing from sealing at the top, which
would have prevented access to the air.
The dialysis tubing had been preserved in 0.1% sodium azide solution, which was toxic to
the larvae in preliminary tests. Therefore, a few days before use each dialysis tube was
detoxified by soaking in distilled water 3 times for 30 minute intervals, after which it was placed
in a 1.0 % sodium benzoate solution to prevent decay. On the day of the experiment, each tube
was once again soaked 3 times for 30 min to remove the sodium benzoate, and then closed off at
the bottom by folding and then sealing with a plastic-coated wire garbage tie.
To the center axil of 10 of the 20 plants, 30 4th instar Wyeomyia spp. were added to 100
mL of sieved bromeliad water containing 0.4 g of chopped oak leaves. One hundred militers of
sieved bromeliad water and 0.4 g of chopped oak leaves were added to the center axil of the ten
control plants. Then each plant was randomly assigned to one of five cages where their bases
were secured in a small plastic container and held for 48 h at 260C with a 12:12 h light dark
cycle and approximately 80% relative humidity.
The plants were checked every 12 h for leaking, which was observed in 10 of the twenty
plants, a loss of around 30 mL by 24 h. At 24 h, more sieved bromeliad water was added to the
leaky plants corresponding to the amount of water loss detected in the plastic container holding
Ammonia levels and pH of the bromeliad water were measured before and after the
experiment. Ammonia concentration was recorded using the Fisher Scientific Accumet Portable
AP63 pH/mV/ion meter using an ammonia probe, and pH was recorded using a Corning pH-20
After 48 h, the water in each bromeliad was removed by a pipette, and rinsed into a metal
rearing pan with a spray bottle containing tap water. The number of surviving A. albopictus and
the instars of each A. albopictus, determined by head capsule widths, were recorded for each
replicate. An ANOVA was performed in SAS (2002) to determine whether the mean
development stage or survivorship ofA. albopictus differed between experimental (30
Wyeomyia) and control (no Wyeomyia) plants.
Surface Area to Volume Ratio Encounter Competition Experiment
Four different sized artificial bromeliads were built with circumferences of 4.5, 5.5, 7.0,
and 9.0 cm at the 75 mL water level, representing surface area to volume ratios of 1.48, 1.82,
2.48, and 3.83 cm2/mL, respectively. The artificial bromeliads were constructed from 8.5 by 11
inch weatherproof map paper (IGage, Mapping Corporation, Salt Lake City UT), and Perfect
Glue No. 1 (Liquid Nails, Cleveland, OH) was used as an adhesive on the outer parts of the
plant. The inner parts of the plant were water protected and sealed with non-toxic, waterproof,
100% silicone aquarium tank sealant (All-Glass Aquarium, Franklin, WI).
Twenty-five 1st instar A. albopictus, 0.5 g oak leaves, and 75 mL of sieved bromeliad water
were added to 15 control bromeliads of each size. The other 15 of each size represented the
experimental group and differed from the controls by the addition of 30 4th instar Wyeomyia spp..
Each replicate was placed into a plastic container of appropriate size and put onto one of three
randomly chosen shelves in a climate controlled rearing room kept at 260C with a 12:12 light
dark cycle and approximately 80% humidity. On the day of setting, each artificial bromeliad
was marked with a pencil to indicate the 75 mL waterline. To compensate for evaporation, and
the differences in evaporation based on the varied surface area to volume ratios among
treatments, on the 4th and 8th day of the experiment each plant was refilled with sieved bromeliad
water to its original 75 mL water level. After 10 days the experiment was terminated, and all of
the A. albopictus in each replicate were counted, and their instars were recorded based on head
Using PROC GLM in SAS (2002) an ANOVA was performed to detect significant
variation among the bromeliad sizes and between the presence and absence of Wyeomyia spp. in
the development rate (average instar) and survivorship (percent alive) of A. albopictus
Habitat Complexity Experiment
The habitat complexity experiment used the same artificial bromeliads described in the
surface area to volume ratio experiment. In this experiment, the smallest artificial bromeliad
with a circumference of 4.5 cm and a surface area to volume ratio of 1.48:1 was excluded
because it was too narrow for the habitat complexity manipulation and still allow for larval
access to the air. In the other three sizes, 20 1st instar A. albopictus, 20 4th instar Wyeomyia spp.,
0.5 g oak leaves, and 75 mL of sieved bromeliad water were added to 30 bromeliads of each.
Two different types of 3 cm2 Squares were constructed from weatherproof paper to add
habitat complexity to the experiment. The first type of paper square, which represents the low
habitat complexity treatment, remained two dimensional (Fig. 2-1). Three of these squares were
added to half of each of the three bromeliad sizes. The second type of paper square, which
represents high habitat complexity, was cut and folded to make a three dimensional structure
(Fig. 2-2). When added to the cone, the low complexity square remained at the top or along the
sides, while the high complexity square made a lattice-like structure within (Fig. 2-2).
After 5 days, the experiment was terminated, and the numbers and developmental stage
of the remaining 1st instar A. albopictus larvae were recorded as described in previous
experiments. An ANOVA was performed in SAS (2002) with survivorship and development
time ofA. albopictus as dependent variables and surface to volume ratio of the artificial
bromeliad (n=3) and the internal complexity (n=2) of the habitat as the independent variables.
Bonferroni-adjusted multivariate pairwise means comparisons followed detection of significant
effects by ANOVA.
Resource Dependent Location of A. albopictus and Wyeomyia spp.
To determine whether foraging behavior ofA. albopictus and Wyeomyia spp. changed in
the presence of each other, or in different levels of food resources, an experiment was conducted
in which three treatments within plastic cups received 0. 1 g of dried chopped oak leaves (Q.
virgininia) as prepared in previous experiments, and the other three treatments received 0.5 g of
the same leaves. The leaves were located at the bottom of all containers. As described in prior
experiments, 75 mL sieved bromeliad water was also added to each container. Each of the food
levels had one treatment with 10 1st instar A. albopictus and 10 4th instar Wyeomyia spp., one
treatment with only 10 1st instar A. albopictus, and one treatment with only 4th instar Wyeomyia
spp.. Each of these treatments was replicated two times. All Wyeomyia spp. were collected as
larvae from bromeliads in Vero Beach and Fort Pierce, FL. All A. albopictus came from a
colony described in the previous experiments.
An instantaneous scan census (Martin and Bateson 1986) was implemented every hour for
7 h starting after a 3 h acclimation period. Treatments with only one species had fewer
individuals, so behavioral observations were usually finished in approximately 15 seconds. The
treatments with both species had more individuals to examine, so those observations often took
as long as 45 seconds. The two replicates for each treatment were sampled consecutively before
moving onto the next treatment, and the treatment observed first was changed at each
ob servati on. To make observations of 1st instar A. albopictus easier, and to prevent disturbing
the larvae with light at every observation, treatments were illuminated by a 60 watt desk lamp for
the entirety of the experiment.
Because the changes in location within the container were the focus of this experiment, the
locations of each Wyeomyia spp. and each A. albopictus at each time were coded into one of four
categories, predetermined from preliminary observations: 1) at the middle of the cup (1cm from
top to 1cm from bottom); 2) on the bottom of cup (bottom to 1 cm from bottom); 3) at the top of
cup (water surface to 1 cm below the water surface with or without siphon extended for
breathing) or 4) wandering (swimming without noticeable foraging behavior). Owing to the
small size of 1st instar A. albopictus, whether the larva was resting, filter feeding or browsing
was not recorded because the differences in these behaviors often could not be determined. For
instance, during preliminary examinations there were multiple occasions when A. albopictus was
nearby a surface, so it appeared to be browsing on that surface, but after closer examination at a
higher magnification, the larva was found to be filter feeding close to the surface, not browsing
At the end of the experiment, the raw data in each observation category were converted to
a proportion based on the number of individuals of each species that showed the categorical
behavior at the observation time. The proportional data were arcsine square root transformed
and a multivariate analysis of variance (MANOVA) was performed with the effects of species,
food, alone (whether with or without the other species), time, replicate, and the interactions of
species x treatment, species x food, and treatment and food. Further analysis of variance
(ANOVA) analyses with subsequent Tukey post-hoc tests differentiated the effects found to be
significant in the MANOVA.
There was no significant difference (F1,19=1.03, P=0.32) between the average instars
(development rate) of A. albopictus in dialysis bags with or without Wyeongia spp. outside the
bag (Fig. 2-3). There was also no significant difference (F1,19=2.94, P=0.10) in the arcsine
square root transformed survivorship between the treatment and the control groups (Fig. 2-4).
The mean values of ammonia, and pH showed that there was little change in these
concentrations between pre- and post-experiment (pH: tl9=1.07, P=0.30; ammonia: tl9=1.34,
P=0.20) (Table 2-1).
Surface Area to Volume Ratio Experiment
The A. albopictus in the treatments with Wyeongia spp. were found to have a significantly
lower average in star (LS mean =2.72 & 0.07 SE) as compared to the LS mean (3.96 & 0.08 SE)
with Wyeongia spp. (Fig. 2-5). There was no significant effect of bromeliad size on average
instar, but there was a significant interaction between the presence of Wyeongia spp. and size
(Table 2-2). From smallest to largest size, the average instars (mean and standard error) at each
surface area to volume ratio were 3.29 & 0.17, 3.54 & 0.16, 3.32 & 0.16, 3.22 & 0.12 respectively.
In treatments without Wyeomyia spp., A. albopictus were found to have a significantly
lower survival (Table 2-3) (0.83 & 0.02 (mean & SE)) compared to the control (0.89 & 0.02
(mean & SE)). The largest three surface area to volume ratios contributed the most to the
significant Wyeomyia spp. effect on survivorship (Fig 2-6). A Tukey post hoc test determined
that the artificial bromeliad with the 6 cm diameter, which was the second to smallest diameter,
had a higher mortality rate than the other three sizes, but was only significantly different from
the smallest size (Fig. 2-6).
Habitat Complexity Experiment
There was no difference in the survivorship of A. albopictus between high and low
complexity treatments (F1,89=0.57, P=0.45) or surface area to volume ratios (F2,89=0.74, P=0.48).
There was no significant difference in the development rate ofA. albopictus among surface area
to volume ratios (F2,89=1.29, P=0.28) but A. albopictus developed faster in high complexity
treatments compared to the low complexity treatments (F1,89=30.02, P<0.01) (Fig 2-7) There
was no significant interaction between surface area to volume ratio and habitat complexity on
developmental rate (F2,89=2.54, P=0.08), or on survivorship (F2,89=0.89, P=0.42).
Resource-Dependent Foraging Location
MANOVA indicated significant effects of species, presence of other species, and
resource levels (Table 2-4). In general, Wyeomyia spp. foraged on the sides and on the bottom of
containers more than A. albopictus (Figs. 2-8 and 2-9). Aedes albopictus were found at the top
of the container more frequently and wandered more than Wyeomyia spp..
There were some significant differences between the locations of the mosquito species at
the different food concentrations (Table 2-5 and 2-6). At high food concentrations Wyeomyia
spp. allocated more time at the water surface (Fig 2-11). Aedes. albopictus also spent more time
at the water surface at high resource concentrations (Fig. 2-10). At low food concentrations,
Wyeonyia spp. allocated more time on the bottom of the container, while A. albopictus occurred
on the sides on the container more frequently (Figs 2-10 and 2-11). There was significant
variation in the location of A. albopictus attributable to the presence of the other species and
food level (Table 2-5).
In the presence of Wyeonzyia spp., A. albopictus spent more time on the top of the
container and less at the bottom or on the sides of the containers (Fig. 2-8). There were
significant interactions between the presence of the other species (alone effect) and species, and
the presence of the other species, and the level of food resources (Table 2-4). Thus, there was a
difference in larval behavior of A. albopictus in the presence of Wyeonzyia, and the level of food
resources also influenced the relative location of these two species.
The chemical interference experiment provided no evidence that waste products from
Wyeonyia spp. larvae interfere with the growth or survivorship of A. albopictus. The buildup of
waste products in experiments that exhibited chemical interference, such as among A. aegypti
(Bedhomme et al. 2005, Dye 1984), may be less likely to occur in the bromeliad tank and axil
habitat. Billbergia pyranzida~lis is both an epiphytic and ground dwelling plant (Frank et al.
1988). While there are no specific reports on the nitrogen absorption abilities of B.
pyra~nidalis, in general, epiphytic bromeliads use dissolved nitrogen held in their tank water as
their main nitrogen source (Benzing et al. 1976, Raven 1988). Epiphytic bromeliads also
assimilate most of their nitrogen from ammonia (Endres and Mercier 2001), which is a common
excretory product of mosquitoes (Clements 1999). By the end of the experiment, there was little
change in the ammonium concentration or pH of the bromeliad water (Table 2-1 ), even though
30, 4th instar Wyeonyia spp., and 25 A. albopictus had been added to the tank for 48 hours.
Because B. pyra~nidalis is sometimes epiphytic, it is possible that the plant absorbed the
ammonia excreted by the mosquitoes, thus reducing the probability of chemical interference
from nitrogenous wastes in this phytotelm. On the other hand, at 24 hours as much as 30 mL of
water was added to some of the plants that had leaked water. This addition of water could have
reduced or eliminated the effects of chemical interference of Wyeomyia spp. on A. albopictus in
The dialysis bag used in the chemical interference experiment physically separated A.
albopictus from Wyeomyia spp. Thus, other potential forms of interspecific competition, such as
encounter, resource, or space competition were eliminated. In some other experiments of this
chapter, A. albopictus were in direct contact with Wyeomyia spp., which resulted in a significant
reduction in growth ofA. albopictus. In the chemical interference experiment, no significant
effects were detected, suggesting that encounter, resource or space competition plays an
important role in the reduction of growth and survivorship seen in the other experiments.
Previous experiments where surface area to volume ratios were manipulated in order to
influence the impact of encounter competition have often been unsuccessful. Broadie and
Bradshaw (1991) altered the surface area to volume ratio of treehole microcosms, but only the
density ofA. sierrensis had an effect on pupation success, larval development time, or pupal
weight. Anholt (1990) altered the surface area available to damselfly larvae in cages, which had
no effect on density dependent decreases in growth and survival. Damselfly larvae exhibited a
behavioral response to the increase of structural complexity inside cages, but habitat complexity
did not change the decrease in growth, survival time, and size seen at high densities (Anholt
In the current study, while the surface area to volume ratio of the plant had no effect on
the development rate of A. albopictus, there was a significant variation in survivorship of this
species among the different surface area to volume ratios. Mean survivorship in size 2 differed
significantly from size 1. The biological significance of this difference between these two sizes is
Another unexpected result of the surface area to volume ratio experiment was the reduced
survivorship ofA. albopictus in the absence of Wyeomyia spp.. The three largest surface area to
volume ratios appeared to contribute to most of this reduction in survivorship, although the
interaction between plant size and Wyeomyia spp. was not significant (Table 2-3).
While the results of the surface area to volume ratio experiment do not explain the effects
of Wyeomyia spp. on A. albopictus, increased habitat complexity in this experiment reduced the
development time ofA. albopictus in the presence of Wyeomyia spp.. In the habitat complexity
experiment the addition of waterproof paper increased the foraging surface area This increase
provided additional surface area to the developing A. albopictus, a factor which was not included
in other competition experiments in this thesis. In this experiment, while more foraging area was
added to the treatments, the total foraging area between treatments was kept the same, because in
both treatments the size of the paper added, and the number of pieces were kept constant, the
paper was folded only to increase complexity. By the end of the experiment though, many of the
low complexity treatment squares had become appressed to the sides of the artificial bromeliad,
thus resulting in a decrease in the foraging surface area between the two habitat complexities.
Therefore, habitat complexity may have decreased the development time ofA. albopictus in the
presence of Wyeomyia spp., but the difference in foraging surface area between the habitat
complexities may have contributed to this decrease in development time. If this effect is due to
habitat complexity, its relevance under Hield conditions is uncertain. When Wyeomyia are
present in bromeliads, which is the case in most of southern Florida (O'Meara et al. 1995b), then
increased complexity could decrease the development time of A. albopictus. Whether increased
habitat complexity affects the developmental rate ofA. albopictus in the absence of Wyeomyia
spp. remains uncertain as no treatments without Wyeomyia were used in this experiment.
The results of the behavior experiment indicated that Wyeomyia spp. spent most of the
time foraging on the bottom of the container, especially during periods of low food
concentrations. When in contact with Wyeomyia spp., A. albopictus spent significantly less time
on the bottom of the container, and stayed the furthest away from Wyeomyia spp. by spending
significantly more time at the top of the container irrespective of food concentration. Such
behavior ofA. albopictus could indicate avoidance of Wyeomyia spp.
These experiments do not support chemical interference as a likely competitive
mechanism to explain negative effects of Wyeomyia spp. on A. albopictus. From the habitat
complexity experiment, encounter competition is the likely explanation for the increase in
development time ofA. albopictus in the presence of Wyeomyia spp., but, as mentioned before,
increased foraging area may have also contributed to this effect. Like other surface area to
volume ratio habitat experiments performed with mosquitoes, this surface area to volume ratio
experiment failed to explain the competitive interactions between A. albopictus and Wyeomyia
spp.. The behavioral experiment provided some indication that A. albopictus larvae change
location within the container in the presence of Wyeomyia spp. These changes in location could
be a result of encounter competition causing A. albopictus to change their location to avoid
encounters with Wyeomyia spp. This avoidance of Wyeomyia spp. may contribute to the
decrease in development rate observed among A. albopictus in the presence of Wyeomyia spp..
Table 2-1. Means and standard errors for the concentrations of ammonia and pH pre- and post-
pH (ppm) (mean & SE) Ammonia (mean & SE)
Pre-experiment 6.21 & 0.02 0.22 & 0.01
Post experiment 6.19 & 0.02 0.21 & 0.01
Paired t-tests showed no significant differences (alpha= 0.05) between pre- and post-experiment
Table 2-2. Analysis of variance for A. albopictus average instar based on the presence or absence
of Wyeomyia spp. and the surface:volume ratio of the artificial plant and the
interaction of these two variables.
Wyeomyia x Size
Table 2-3. Analysis of variance for A. albopictus average survival based on the presence or
absence of Wyeomyia spp., and the surface area:volume ratio of the plant (size) and
the interaction of these two variables.
F Df P-value
Wyeomyia 6.48 1 0.01
Bromeliad Size 5.41 3 <0.01
Wyeomyia x Size 1.56 3 0.20
Table 2-4. Multivariate analysis of variance table for feeding location.
Effect Wilks' Lambda Df P-value
Species x Alone
Species x Food
Alone x Food
The species effect refers to the location of either A. albopictus or Wyeomyia spp.. Alone refers to
whether or not each species is alone or in the presence of the other species. The food category
variables were either 0. 1 g leaves (low treatment) or 0.5 g leaves (high treatment). Time is the
hour of observation from the start of the experiment.
Table 2-5. Analysis of variance for the effects of food level presence or absence of Wyeomyia
spp. on the location of A. albopictus.
F Df P-value
Combo High Food 15.40 3,63 <0.01
A. albopictus only High Food 3.49 3,63 0.02
Combo Low Food 12.24 3,63 <0.01
A. albopictus only Low Food 1.12 3,63 0.35
Combo indicates the treatments that had both Wyeomyia spp. and A. albopictus.
Table 2-6. Analysis of variance for the effects of food level and presence or absence of A.
albopictus on the location of Wyeomyia spp.
F Df P-value
Combo High Food 14.55 3,63 <0.01
A. albopictus only High Food 33.87 3,63 <0.01
Combo Low Food 39.58 3,63 <0.01
A. albopictus only Low Food 15.58 3,63 <0.01
Combo indicates the treatments that had both Wyeomyia spp. and A. albopictus.
Figure 2-1. Construction of the squares for the habitat complexity experiment. A) the low
complexity square is unaltered before introduction to the bromeliad. B) the high
complexity square was cut along the lines indicated inside the square.
A BB C
Figure 2-2. The folded shape of the high complexity and low complexity treatments and the
orientation of the squares within the cones. A) in the low complexity experiment the
squares floated on the top, or laid on the sides of the cone. C) in the high complexity
cone, the altered squares create a lattice-like formation on the inside of the cone. B)
The diagram in the center shows the shape of the altered square after the cutting and
folding described in Fig. 2-1.
~ieomyia Absent o~myia Present
Figure 2-3. Average instars of A. albopictus after 48 h exposure to Wyeomyia spp. through a
dialysis membrane and without Wyeomyia in the control. The error bars are standard
FF)eomyia Ab~sent Wyeomyia Present
Figure 2-4. Mean survivorship of A. albopictus after 48 h exposure to Wyeomyia spp. through a
dialysis membrane and without Wyeomyia spp. in the control. The error bars are
811b b b
< 1.5 -
NoWy Wy NoWy Wy NoWy Wy NoWy Wy
Size 1 Size 2 Size 3 Size 4
Figure 2-5. The average instar + SE ofA. albopictus in the absence or presence of Wyeomyia
spp. in four plant sizes. NoWy refers to the treatments with no Wyeomyia spp., and
Wy refers to treatments with Wyeomyia spp.. The size below these symbols refers to
the surface area to volume ratio in the artificial bromeliad, proceeding from smallest
(1) to largest (4), as quantified in Materials and Methods. Means with common
letters above the bars are not significantly different by (P<0.05).
I I I I
Nos~ Wy NoWy Wy NosWy Wy NO~Cy Wyy
TSize 1 Size 2 Size 3 Size 4
Figure 2-6. The average survivorship + SE ofA. albopictus in the absence or presence of
Wyeone i spp. in four plant sizes. NoWy refers to the treatments with no Wyeonglia
spp., and Wy refers to treatments with Wyeone ia spp.. The size below these symbols
refers to the surface area to volume ratio in the artificial bromeliad, proceeding from
smallest (1) to largest (4), as quantified in Materials and Methods. Mean values
without common letters above the bars are significantly different (P<0.05).
L2 H2 L3 H3 L4 H4
Figure 2-7. The average instar (+SE) of A. albopictus in two levels of habitat complexity and
three levels of surface area to volume ratio. L and H denote the low and high
complexity treatments. Numbers correspond to increasing surface area to volume
ratios, as indicated in Materials and Methods. Mean values without common letters
above the bars are significantly different by a Bonferroni adjusted multivariate
Wyeomyia No Wyoemyia
Location of Mosquet
[ ] Bottom
Figure 2-8. Location ofA. albopictus in the presence and absence of Wyeomyia. The bottom and
top locations are significantly different between Wyeomyia and no Wyeomyia
treatments (alpha<0.05) by a Tukey means comparison. The bars indicate SE.
No A. abdcpictubS A. abdcpictubS
Locatio n of M osquito
Figure 2-9. Location of Wyeomyia spp. in the presence and absence ofA. albopictus. No
locations were significantly different between A. albopicats and no A. albopicats
treatments (alpha<0.05) by a Tukey means comparison. The bars indicate SE.
0 a 0
o- a a:
~0a a SideI
Fiur -1.Loato oA aboitu y od cocnrto adtepee cando absenc sof u~
Figue 2indiLcate igniofic anly diffretu menaogh food concentrations (high pesn and lbeeow)
and treatments (Wyeone ia spp. and No Wyeone ia spp.) for each location (side,
bottom, top and wander) (alpha=0.05).
LU0.4 b b
o7 b b
ff 44Location of M ~osuito
go~ go Wander
Figure 2. 11: Location of Wyeomyia spp. by food concentration and the presence and absence of
A. albopictus. The bars indicate SE. Different letters over the same bar category
indicate significantly different means among the food concentrations (high and low)
and treatments (A. albopictus spp. and No A. albopictus spp.) for each location (side,
bottom, top and wander) (alpha=0.05)
Wyeomyia spp. AS POTENTIAL PREDATORS OF Aedes albopictus IN BROMELIADS IN
In mosquito populations predation can effect the survivorship (Griswold and Lounibos
2006, Blaustein et al. 1995, Fincke et al. 1997), and development time (Alto et al. 2005, Grill
and Juliano 1996) of the prey as predators consume prey or the prey become less active in the
presence of predators. These alterations in survivorship and development time can effect the
population and community structure of both the predator and prey species (Sih 1985).
Intraguild predation is the killing and consumption of competitors that consume similar
resources, which has an immediate energetic benefit and a beneficial reduction in exploitative
competition for the predator (Polis et al. 1989). Age and size dependent intraguild predation is
common throughout all ecological systems, and many predators are cannibalistic on smaller
sized or younger conspecifics (Polis et al. 1989). Often intraguild predation occurs between
competitors with the greatest resource overlap, and may be a result of the increased encounter
rates due to niche overlap (reviewed by Polis et al. 1989).
Mosquitoes that are predominantly browsers and filter feeders, such as Aedes and
Anopheles species (Clements 1999), may become cannibals (Hinman 1934, Reisen and Emory
1976, Koenekoop and Livdahl 1986, Koenraadt et al. 2003) or facultative predators (Edgerly et
al. 1999, Koenraadt et al. 2003) under certain conditions. Koenekoop and Livdahl (1986) and
Edgerly et al. (1999) found that cannibalism in Aedes triseriatus was resource dependent with
lower resource availability leading to higher cannibalism. Koenraadt et al. (2003, 2004) found
no increase in cannibalism or predation by Anopheles gambiae complex mosquitoes related to
resource availability. Increased habitat complexity also reduced predation by A. triseriatus on
congeners, but the opposite effect was seen with A. aegypti (Edgerly et al. 1999). Koenraadt et
al. (2004) also showed that higher densities of anopheline mosquitoes in small spaces increased
cannibalistic and predatory behavior.
Although facultative predation may influence the population dynamics of many Aedes and
Anopheles mosquitoes, W. mitchellii and W. vanduzeei have yet to be examined as potential
predators of A. albopictus in bromeliads. Because predation is an important mechanism shown to
effect the survivorship and development time of mosquito species in other studies, Wyeomyia
spp. larvae from Florida bromeliads were dissected after exposure to A. albopictus to determine
whether these Wyeomyia spp. may be facultative predators ofA. albopictus.
Materials and Methods
Testing for Density Dependent Predation by Wyeomyia spp.
In this experiment, three densities (0, 25, 50) of 1st instar A. albopictus were placed into
plastic containers containing 75 mL sieved bromeliad water and 0.5 g dried (680C for 48 h),
chopped Quercus virginiana leaves. Thirty-six containers were used, allowing 12 replicates of
each density. To each container thirty 4th instar Wyeomyia spp. were added after previously
being kept in tap water without food for 36 h prior to the experiment. The lack of food for
Wyeomyia spp. prior to the experiment was to ensure that all gut contents examined after the
experimentation resulted from the food obtained during the experiment, and not prior feedings.
After 12 h, 6 containers were removed from each density, and all live 4th instar Wyeomyia spp.
larvae were dissected. At 24 h, the remaining containers were removed and again all of the live
4th instar Wyeomyia spp. larvae were dissected. Dead larvae of Wyeomyia spp., and pupae were
Testing for Resource Dependent Predation by Wyeomyia spp.
In a second experiment, two densities of food resources (0. 1 g and 0.5 g) of dried (680C
for 48 h), chopped Q. virginiana leaves were placed into 24 plastic containers with 75 mL sieved
bromeliad water and 25 1st instar A. albopictus. Twenty-five 4th instar Wyeomyia spp. were
added to each container. Twelve hours and 96 h after the start of the experiment, 6 containers at
each food level were removed and their live 4th instar Wyeomyia spp. were dissected. As before,
dead Wyeomyia spp., and pupae were not dissected.
Visual examinations of the behavior of A. albopictus and Wyeomyia spp. were conducted
in plastic containers. Ten 1st instar A. albopictus were added with either (1) 0. 1 g food and no
Wyeomyia spp.; (2) 0.1 g food and 10 4th instar Wyeomyia spp.; (3) 0.5 g food and no Wyeomyia
spp.; or (3) 0.5 g food and 10 4th instar Wyeomyia spp.. There were two replicates of each
combination. The food was dried, chopped Q. virginiana leaves as used in the two previously
described predation experiments. Two cups at each food level were selected, every hour for 10
h, to be examined for signs of predation by Wyeomyia spp. on A. albopictus. Predatory behavior
was classified as chewing, biting, grabbing or actually consuming another organism. Each
examination occurred for 1 min, and since light was necessary for visualization of the 1st instar
A. albopictus, a desk light kept the cups illuminated for the entire 10 h duration of the
No A. albopictus body parts were found in the guts of any of the 941 Wyeomyia spp.
dissected (Table 3-1). All guts of Wyeomyia spp., including the treatment without A. albopictus,
contained brown organic material, but no signs of head capsules, or any other body parts, ofA.
albopictus. There were also no signs of aggression upon A. albopictus by Wyeomyia spp.
throughout the entire duration of the predation observation experiment.
From these experiments, it seems very unlikely that Wyeomyia spp. prey upon A.
albopictus. When field-collected larvae of W. vanduzeei were examined by Fish (1976), no
remains from any invertebrates were found; only organic particulate matter, protozoans and
pollen grains were found in the gut contents. In a study by Broberg and Bradshaw (1995), the
pitcher plant mosquito, Wyeomyia smithii, was found to not exhibit density dependent
cannibalism. Therefore, although other detritivores, such as some Aedes and Anopheles species,
can be cannibalistic or facultative predators, it seems that Wyeomyia spp. of North America do
not exhibit these characteristics, at least under the density and resource conditions examined in
As an adaptation to the variable water and resource levels in a bromeliad, W. vanduzeei can
develop slowly on limited food resources (Frank and Curtis 1977), so the Wyeomyia spp. of
Florida bromeliads may not have adopted a predation strategy, because they have adapted to a
food limited environment. Most Aedes and Anopheles species that are cannibalistic or facultative
predators would starve at food levels that Wyeomyia spp. can subsist upon (Barrera and
As with any lab experiment, the applicability of these results to nature comes into question.
As mentioned in other chapters, bromeliad phytotelmata vary in size, structure, complexity,
water holding capacity, and faunal composition. All of these factors could influence the
expression of predation in Wyeomyia spp.. Variations of bromeliad phytotelm size, structure,
and complexity have been examined to some extent in this study. In this study, whether
conducted within an actual bromeliad or an artificial one, there was no significant reduction in
survivorship of A. albopictus that would indicate predation. In conclusion, if there is predation
by Wyeomyia spp. on A. albopictus it occurs so rarely as to be unimportant in overall
survivorship of A. albopictus.
Table 3-1. Number of Wyeomyia examined in predation experiments.
Density Dependent Resource Dependent
Time 0 Albo. 25 Albo. 50 Albo. Low food high food
12 h 111 123 128 74 111
24 h 117 126 151 n/a n/a
96 h n/a n/a n/a 87 158
This table compiles the number of dissected Wyeomyia spp. from two different experiments.
The time box indicates the number of hours after A. albopictus were added that Wyeomyia were
examined. The next three categories, O Albo., 25 Albo., and 50 Albo., correspond to the first
experiment in which the Wyeomyia were examined in the presence of 0, 25 and 50 1 st instar A.
albopictus. The next two categories, low food and high food, indicate the treatments in the
second experiment in which Wyeomyia spp. were examined after exposure to 0.1 g or 0.5 g of
food in the presence of 25 1 st instar A. albopictus.
DO Wyeomyia spp.. LARVAE INTHIBIT EGG HATCH OF Aedes albopictus?
Eggs of Aedes spp. hatch after submersion in water, and low oxygen levels caused by
nearby microbial growth stimulate hatching (Gjullin et al. 1941, Judson 1960, Fallis and Snow
1983). Most eggs will hatch in response to the initial stimulus, but some hatching is delayed,
awaiting later stimuli (Gillett et al. 1977, Livdahl and Koenekoop 1985, Andreadis 1990).
Suppression of egg hatch by larvae occurs when the larvae consume microbes on and near the
eggs, thus causing a reduction in the local microbial population, and a subsequent increase in
oxygen resulting in inhibition (Gillett et al. 1977, Edgerly and Marvier 1992) There are also
abiotic secondary determinants of hatch inhibition, such as temperature (Mallack et al. 1964),
photoperiod (Horsfall 1956, McHaffey and Harwood 1970, McHaffey 1972, Shroyer and Craig
1980), and variations in wet and dry periods or humidity (Andreadis 1990, Clements 1999),
which will also delay hatch by putting eggs into quiescence or diapause until conditions are
favorable for hatching .
Egg hatch inhibition is hypothesized to benefit larvae by delaying hatch when
competition for resources is high (Livdahl et al. 1984, Livdahl and Edgerly 1987), risk of
predation is high (Koenekoop and Livdahl 1986), or when abiotic conditions are unfavorable for
development (Shroyer and Craig 1980, Clements 1999). For many Aedes species, these
benefits are thought to outweigh the costs of egg hatch inhibition, which increases the time until
reproduction. Increasing the time to reproduction may lead to a reduction in population growth
and may increase mortality in the dormant egg stage of aedine mosquitoes (Livdahl and Edgerly
Because many Aedes mosquitoes exhibit some level of hatch inhibition (Edgerly et al.
1993), it is important to examine this role of competition on Aedes hatch. Within the exotic
bromeliads of southern Florida, larval competition between Wyeomyia spp. and A. albopictus is
thought to influence the mosquito community structure in this system (Lounibos et al. 2003),
although the mechanism for this competition has yet to be determined. Field collections from
exotic bromeliads in southern Florida show that high densities of 4th instar Wyeomyia spp. are
common in exotic bromeliads (Raban unpublished data). Additionally, Wyeomyia larvae
develop slowly, as an adaptation to the variable water and resources levels in a bromeliad (Frank
and Curtis 1977), so it is feasible that these large, later instar larvae are present for long
durations. Also, W. mitchellii and W. vanduzeei are found in the larval stage throughout the
year, while some A. albopictus overwinter as eggs. Thus, even an initial spring A. albopictus
cohort could experience egg hatch inhibition from the presence of Wyeomyia spp. Therefore, a
resource dependent egg inhibition experiment was conducted to determine whether older instar
Wyeomyia spp. inhibit the hatch of A. albopictus eggs and whether inhibition varies with larval
Materials and Methods
Fifty A. albopictus eggs on each of 56 papers inserted individually in plastic cups were
submerged in 75 mL of sieved bromeliad water. All A. albopictus eggs used in the study were
from one oviposition paper, resulting from approximately 3 days of oviposition from a Florida
colony of this species supplemented irregularly with wild mosquitoes. All eggs were counted
under a dissecting microscope to ensure no hatch or damage before the start of the experiment.
Twenty-eight of these cups were randomly assigned to a low food level of 0. 1 g of dried,
chopped, live oak leaves (Q. virginiana), and the remaining 28 were assigned a high food level
of 0.5 g of dried, chopped, live oak leaves. All leaves were dried at 410C for 72 h before the
experiment. At each food level, 20 field collected, 4th instar Wyeomyia spp. were added to 14 of
the cups. The other fourteen cups contained only the 50 A. albopictus eggs and served as
controls. After 48 h the experiment was terminated, and the egg hatch rates were recorded for
each cup. The number of eggs hatched was determined by counting the number ofA. albopictus
larvae present and the number of eggs that hatched on the paper. Whichever hatch number was
higher, number of larvae or the number of eggs hatched on the paper, was considered the hatch
number for that cup. In almost all cases, the number of eggs hatched on the paper exceeded or
equaled the number of larvae found in the cup.
The bleaching technique of Trpis (1970) was used to determine the viability of the
unhatched eggs. Eggs were considered viable if they had a fully developed embryo. Empty eggs
and partial or shriveled embryos were considered inviable and were omitted in the data analysis.
On the other hand, if eggs were not wholly encased and appeared to be crushed or damaged,
these eggs were classified as damaged.
Due to the high number of inviable eggs, an ANOVA was conducted on their incidence to
ensure their uniform spread across treatments. To ensure homogeneity of variances, hatch rates
were arcsine square root transformed before an ANOVA was performed in SAS (2002) with a
subsequent Tukey's post-hoc test. A Kruskal-Wallis test and a Dunnet' s test was performed on
the damaged egg data. Two separate treatments had one replicate each that was omitted from the
analysis because they both had a 0.0% hatch rate with only 2 to 4 viable eggs per replicate.
Overall, 29.0% + 2.4% (mean & SE) of all eggs were inviable, with no difference among
treatments (F3,50= 1.88, P= 0.14). There was a significant difference between the hatch rates of
the treatments (F3,50=6.28, P=0.02), which was attributed to the high food with Wyeomyia spp.
treatment being significantly different than both food treatments without Wyeomyia spp. (Fig. 4-
1). There was a lower hatch rate for the treatments with Wyeomyia spp. as compared to the
treatments without Wyeomyia spp., although this difference was only significant for the higher
food concentration (Fig. 4-1). On the other hand, there was a significant variation (X2=8.06, df=3,
P=0.04) among treatments in the incidence of damaged eggs (Fig. 4-2). Significantly more
damaged eggs were observed in the low food treatment with Wyeomyia spp. compared to both
the food treatments without Wyeomyia spp. (Fig. 4-2).
In this experiment, there was a significantly lower hatch rate ofA. albopictus in the
presence of fourth instar Wyeomyia spp at higher food concentrations. These Eindings are not
overly surprising as previous studies by Edgerly et al. (1993) showed reduced hatch of A.
albopictus in the presence of high densities of larger instar Aedes larvae at 24 h. While the hatch
rate of A. albopictus in the lower resource treatment with Wyeomyia spp. was lower than the
lower resource control, this difference was not significant, indicating that resource level may
influence the egg hatch inhibition in A. albopictus.
Even though in this experiment egg hatch inhibition by Wyeomyia on A. albopictus was
found to influence the hatch rate at high food concentrations, the effects on hatch rate were only
moderate. In this experiment, treatments with Wyeomyia experienced only an approximately 9%
decrease in hatch rate at 48 h, and 80% to 90% of viable eggs still hatched in all treatments.
Therefore, while there may be egg hatch inhibition, if the hatch rates seen in this experiment are
indicative of natural field hatch rates, then the influence of egg hatch inhibition on A. albopictus
development is not pronounced and does not explain the decrease in development rate of A.
albopictus in the presence of Wyeomyia spp. seen in the surface area to volume ratio experiment,
and the larval experiment in Lounibos et al. (2003).
The reduced effects of egg hatch inhibition on A. albopictus are not surprising, as Edgerly
et al. (1993) found that at higher larval densities A. albopictus egg hatch was inhibited, but at
reduced rates as compared to other Aedes spp. mosquitoes tested. So, in general it is possible
that A. albopictus does not experience pronounced egg hatch inhibition. Also, this experiment
was allowed to run for 48 h, while the Edgerly et al. (1993) experiment was terminated at 24 h,
so it is possible that most pronounced egg hatch inhibition occurs in the first 24 h, which was
missed in a 48 h experiment.
Livdahl et al. (1984) and Livdahl and Edgerly (1987) found that Aedes triseriatus hatch
rates are reduced by large densities of late instar conspecifics. These authors hypothesized that
large 4th instar larvae inhibit hatch through their elevated grazing intensity, which reduces
oxygen-depleting microbes. The significantly larger amount of egg damage seen in the presence
of Wyeongia spp. in this experiment may be a result of browsing by the late instar Wyeongia
spp. on the eggs ofA. albopictus. Because the percentage of eggs damaged was highest in the
low resource group (but not significantly different from high resources) with Wyeongia spp., it
suggests that with fewer leaf resources to graze upon, Wyeomyia spp. may have increased their
grazing upon A. albopictus eggs. The maj ority of egg damage probably occurred after hatch or
did not effect hatch rate, because the high resource with Wyeongia spp. treatment had lower
hatch rate than the lower resource with Wyeongia treatment, but the lower resource treatment
with Wyeongia spp. had higher egg damage. Likewise, the lower resource with Wyeongia spp.
treatment did not have a significantly different hatch rate from either of the controls at each
resource level, but it did have significantly more egg damage. Thus, there seems to be no strong
relationship between hatch rate and egg damage.
In nature, egg hatch inhibition ofA. albopictus within bromeliads could differ from the
laboratory results. In the typical bromeliad habitat many other organisms could also be
influencing the bacterial content of the area around the eggs and, thus, the hatch rate of A.
albopictus eggs. Many other bacteria- and detritus-consuming macroinvertebrates are known to
inhabit bromeliads (Fish 1976, Frank 1983).
Additionally, the bromeliad itself can influence the chemical composition of its contents
based on its physiological needs (see bromeliad as a habitat section in Chapter 1 of this thesis).
For instance, Fallis and Snow (1983) determined that an increase in water nitrogen induced hatch
in Aedes punctor, and nitrogen is one of the main chemicals that is absorbed by bromeliads
(Benzing 2000), and thus may influence the hatch ofA. albopictus. Fallis and Snow (1983) also
found that the change in oxygen content, and not the concentration of oxygen was the hatching
trigger for A. punctor, which may also be regulated by the bromeliad. So, although Wyeomyia
spp. may inhibit the hatch ofA. albopictus eggs at high resource levels, it is possible that other
organisms or processes within the bromeliad can alter egg hatch inhibition in A. albopictus.
With more macroinvertebrates in the aquatic community, there could be more grazing
and an increase in localized oxygen levels near the eggs, and thus more inhibition. In contrast,
Edgerly and Marvier (1992) hypothesized that at a certain density, the number of organisms
surrounding eggs can be great enough to deplete the overall oxygen levels in the water, and thus
stimulate egg hatch. Although as mentioned before, Edgerly et al. (1993) found that at high
densities A. albopictus had a low level of hatch inhibition compared to congeners, so higher
densities of mosquitoes may have little effect on the overall hatch ofA. albopictus.
In this experiment the eggs of A. albopictus were submerged below the water line in a
cluster on the side of the container. In the location study from Chapter 2 of this thesis
Wyeomyia spp. larvae were located more often at the bottom and at the top of the container and
less on the sides. If A. albopictus does not oviposit its eggs on the inner axil walls of bromeliads,
then this experiment may misrepresent the influence of Wyeomyia spp. on egg hatch inhibition
of A. albopictus. There is no evidence that A. albopictus lays its eggs on the sides of the
bromeliads, nor is it probable that their eggs are laid in one cluster as in this experiment. Further
investigation into the location and distribution of oviposited eggs of A. albopictus within
bromeliads is needed to determine the relevance of this experiment to natural conditions.
High Food High Food Low Food Low Food
No Wvy Wy No WvyW
Figure 4-1. Proportion hatch of viable eggs + SE in each treatment. High Food refers to high
food treatments, and Low food refers to low food treatments. No Wy indicates
treatments with no Wyeomyia spp. and Wy indicates treatments with Wyeomyia spp..
Treatments with the same letter above the SE bar are not significantly different based
on a Tukey test.
HihFodHg Fo owFo LwFo
tramns an o odrfrst o odteamns oW ndctstetet
Tiue42 rpretmonents wihthamaed letter trabove theS a areh noot sinfiatliferen basedhf
on a Dunnett' s Test.
FIELD STUDIES ON Aedes albopictus AND Wyeomyia spp. INT EXOTIC BROMELIADS OF
In a study conducted by O'Meara et al. (1995b), Wyeomyia spp. were the most abundant
mosquitoes in bromeliads throughout southern Florida, but in northern Florida where W.
vanduzeei and W. mitchelii are absent, A. albopictus was the most common mosquito species in
bromeliads. In the southern sites with Wyeomyia spp., A. albopictus was often found in great
numbers in nearby artificial containers. For example, when vases were placed near bromeliads
in Vero Beach, Florida, A. albopictus larvae were subsequently collected from all the vases but
from only 40 percent of the bromeliads. In the study of Lounibos et al. (2003), A. albopictus was
also found in bromeliads in greater numbers in the absence of Wyeomyia spp..
Wyeomyia spp. vary in larval abundance and ovipositional preference based on the species
of bromeliad. In collections by O'Meara et al. (1995b), Wyeomyia spp. were less abundant in
Neoregelia spectabilis than in Aechmea fasciata. Frank and O'Meara (1985) also found that W.
vanduzeei showed an oviposition preference for the native bromeliad Tillandsia~~ll1~~~~111~~~ utriculata over
another native bromeliad, Catopsis berteroniana.
The purpose of the current study was to further explore the differences in abundance and
distribution of Wyeomyia spp. and A. albopictus within bromeliads based on the location within
the plant axill or central tank), by bromeliad species, and by macrohabitat. Neoregelia
spectabilis and B. pyramnidalis are two exotic bromeliad species commonly featured in the
residential landscaping of many homes in the cities of Vero Beach and Fort Pierce, Florida.
These two plants were chosen for further study because they were common and distinctly
different in size, color, and shape, but are often grown in the same location. In the current study,
N. spectabilis and B. pyra~nidalis were sampled to compare the abundances of Wyeonzyia spp.
and A. albopictus in these two bromeliad species. Due to the differences in their physical
characteristics these bromeliads may have different abundances of mosquitoes. These samples
were further divided by location within the plant, i.e. lateral axil or central axil, to determine
whether the abundance of Wyeonzyia spp. or A. albopictus varied by location within the plant. It
is also hypothesized that the abundance of mosquitoes will not vary by location within the
bromeliad, because bromeliad studies by Frank and Curtis (1977) have indicated that during
rainfall eggs may be washed to other locations within the plant. Thus, if rainfall occurs
frequently enough there should be a rather homogeneous distribution of larvae throughout the
plant. To further describe habitat preferences of bromeliad-inhabiting mosquitoes, mosquitoes
were collected from the Hield to determine whether mosquito abundance varied between two
canopy types, mainly oak tree and palm tree. Frank and O'Meara (1985) determined that W.
naitchellii preferred shady habitats and, thus, different canopy types could also provide shade
variations that influence the abundance of bromeliad mosquitoes.
Materials and Methods
Differences in the Density of A. albopictus and Wyeomyia spp. within N. spectabilis and B.
The aquatic contents of B. pyra~nidalis and N. spectabilis bromeliads were collected with a
meat baster monthly from September 2005 to July 2006 at eight sites in Vero Beach and Fort
Pierce, Florida, and from two sites in Orlando, Florida (Fig 5-1). On a few occasions plants in
Tampa Bay and Washington Oaks Park were also sampled (Fig 5-1). At each site, Hyve to
twenty-Hyve plants of each species were sampled every 4 to 8 weeks. To avoid taking repeated
samples from the same plants, at each sampling the approximate locations of the samples were
recorded, although repeated sampling may have occurred at two of the smaller Vero Beach, FL
sites due to a reduced number of plants. The sites where re-sampling of plants was suspected
were omitted from all analyses.
From each plant two samples were taken by removing all extractable fluid contents with
the baster. One sample consisted of the contents of the central axil of the plant, while a separate
sample was taken of the lateral axil contents. The central axil sample was extracted from the
area of water held in the center of the plant (photo A in Fig. 2.2), and the lateral axil sample was
extracted from the spaces created by the lateral axils of the plant (photo B in Fig. 2.2).
Due to the fact that B. pyramnidalis with inflorescences held very little, if any, water, only
B. pyramnidalis that were not flowering were sampled in this study. Conversely, since N.
spectabilis holds very little if any water in its central tank when it is not flowering, only
flowering N. spectabilis were sampled in this study. The species and development stage of each
mosquito immature collected were recorded. To aid in the identification of 1st instar larvae, all
samples were initially examined in trays under a dissecting scope.
Studies by Frank et al. (1977) have shown a correlation between bromeliad size and the
number of mosquito larvae and pupae. After February 2006, in the current study the amount of
water in each bromeliad was also measured to determine whether a correlation exists between the
amount of water in a bromeliad and the abundance of either species of mosquito and their total.
Thus, the analysis of the data from these collections is broken into two categories 1) all
collections, with analyses being based on total mosquitoes per plant, and 2) collections after
February 2006, which were analyzed by numbers of mosquitoes per volume of water. ANOVA
calculated from type III sums of squares was applied to detect significant variations in the
dependent variables densityy of mosquitoes per mL and per plant) in relation to the independent
variables (bromeliad species). Means comparisons were conducted with Tukey's tests. The
density means (mos/mL and mos/plant) were also compared by a paired t-test in relation to the
location within the plant (lateral vs. central axil). A repeated measures ANOVA, with
subsequent means comparisons by Tukey's tests, was applied to detect significant variations in
the dependent density variables by the month of collection. Correlations were performed using
Spearman' s p between the volume of water in bromeliads and the numbers of A. albopictus and
Wyeomyia spp. found within.
Canopy Effects on Density
A study was also conducted on the differences in density of Wyeomyia spp. and A.
albopictus in bromeliads based on the surrounding habitat. Preliminary observations during
collections showed that there may be fewer Wyeomyia spp. and A. albopictus in bromeliads with
palm trees overhead versus in bromeliads shaded by oak trees. These habitats differ in amount
of sun exposure, leaf litter, and throughfall input into the bromeliads below.
The largest sample site in Vero Beach, FL (VB-5) contained four clumps of interspersed N.
spectabilis and B. pyramnidalis under palm trees and seven clumps of interspersed N. spectabilis
and B. pyramnidalis under oak trees. In both habitats B. pyramnidalis was the more common of the
two bromeliad species. Once in April 2006 and once in June 2006 two samples, one from the
central axil, and one from the lateral axil, of fifteen B. pyramnidalis and nine N. spectabilis were
taken from under either an oak tree or a palm tree. In order to avoid sampling the same plants
twice, the location of the first collection was different from the location of the second collection.
The mosquitoes collected were counted and recorded as described in the previous section.
Significant effects of canopy type on the numbers of Wyeomyia spp. in bromeliads were tested
by nested ANOVA with bromeliad species (B. pyramnidalis vs. N. spectabilis) nested within
canopy type (Oak vs. Palm). Aedes albopictus was omitted from ANOVA analysis as no larvae
of this species were found in the oak location.
Both mosquitoes per plant (mos/plant) and per unit volume (mos/mL) varied significantly
between plant species (Table 5-1 and Table 5-2). Aedes albopictus had a higher density
(mos/mL and mos/plant) in N. spectabilis, but Wyeomyia spp. was found in significantly higher
densities in B. pyramnidalis (Tables 5-3 and 5-4). Aedes albopictus had higher densities
(mos/plant) in the central axils of N. spectabilis (t217= 2.22, P=0.03), but showed no difference in
density between the central and lateral axils in B. pyramnidalis (t242= 0.86, P=0.39) (Table 5-4).
Aedes albopictus also showed no difference in density (mos/mL) between the central and lateral
axils in either N. spectabilis (tl72= -0.70, P= 0.49) or B. pyramnidalis (tl93= 1.00, P= 0.32).
Wyeomyia spp. showed no differences in densities (mos/plant) between the lateral and central
axils of B. pyramnidalis (t242= 1.38, P= 0. 17) or N. spectabilis (t217= 1.60, P= 0.11). However,
Wyeomyia spp. did have higher mean densities (mos/mL) in the lateral axils of both N.
spectabilis (tl72= 2.27, P= 0.02) and B. pyramnidalis (tl93= 2.52, P= 0.01) (Table 5-4).
The densities (mos/plant) of Wyeomyia spp. and A. albopictus varied by the month of
collection (W.- F6= 8.74, P<0.01 and A. albo.- F6= 12.36, P<0.01) and by the site of collection
(W.- Fs= 10. 18, P<0.01and A. albo.- Fs= 2. 15, P=0.03). The density of A. albopictus in B.
pyramnidalis was relatively constant throughout the year, but in N. spectabilis the density at each
site of A. albopictus increased in September and in April through July (Fig. 5-3). In B.
pyramnidalis the density of Wyeomyia spp. increased in September, decreased in November
through March, and increased in April and in July. Aedes albopictus also experienced similar
changes in density (mos/mL) over time, but Wyeomyia spp. density (mos/mL) did not differ
significantly among months (Tables 5-5, 5-6, Figs. 5-4, 5-5,).
There were also distinct differences in the water holding capacity between the plants
(Table 5-7). More water was extracted on average for N. spectabilis than B. pyramnidalis (Table
5-7). When the data were analyzed separately by the location within the plant there was
significantly more water in N. spectabilis than B. pyramnidalis in both lateral and central axils
(Table 5-7). There were positive correlations between the number of Wyeomyia spp. (p400= 0.12;
P=0.01), the number of A. albopictus (p400= 0.26; P<0.01), and the total mosquitoes (p400= 0.18
P<0.01) and the amount of water extracted from the bromeliad.
Wyeomyia spp. differed in densities between the palm and oak sites (Table 5-8), and by
plant species in the palm and oak sites (Table 5-8). Wyeomyia spp. were less abundant in the
palm sites, and in both the oak and palm sites Wyeomyia spp. was found in a significantly greater
densities in B. pyramnidalis plants (Table 5-9).
In the current surveillance Wyeomyia spp. immatures were denser in B. pyramnidalis than in
N. spectabilis. O'Meara et al. (1995b) showed that Wyeomyia spp. were more abundant in
Aechmea. fasciata than in N. spectabilis. On the other hand, A. albopictus had much higher
mean densities in N. spectabilis than in B. pyramnidalis. The differences in abundances in each
plant could indicate that A. albopictus is avoiding contact with Wyeomyia spp. by ovipositing in
bromeliads not occupied by Wyeomyia spp., but in an experiment conducted by Lounibos et al.
(2003), A. albopictus showed no difference in ovipositional preference between B. pyramnidalis
with Wyeomyia spp. present or absent. The Lounibos et al. (2003) experiment indicates that
ovipositing A. albopictus are not responding to cues from Wyeomyia spp. larvae. The
Washington Oaks site had the greatest abundance of A. albopictus, and the lowest abundance of
Wyeomyia spp. The studies by O'Meara et al. (1995b) and Lounibos et al. (2003) indicate that
more northern sites like Washington Oaks have more A. albopictus in their bromeliads than in
the southern sites. These two studies also indicated that lower abundances of Wyeomyia spp.
occurred in northern Florida, which was likely due to the intolerance of Wyeomyia spp. to
temperatures below freezing, or to the rarity of native bromeliad phytotlemata in northern
There were differences in the densities of Wyeonyia spp. and A. albopictus by location
within the plant. Wyeonzyia spp. had higher densities (mos/mL) in the lateral axils of B.
pyra~nidalis and N. spectabilis, so either Wyeonzyia spp. oviposits more frequently in lateral
axils, or if Wyeonzyia spp. oviposit more eggs into the central axil, then some eggs or larvae are
washed by rainfall into the lateral axils. Frank et al. (1976) observed that female W. vandueezi
oviposit more frequently into the central axils of 7: utriculata, but were still found in the lateral
axils of that bromeliad species. In a Frank and Curtis (1977) study, W. vandueezi were found to
be easily washed out of the central tank of the bromeliad during rainfall. Therefore, it is
unknown whether the lateral axils and central axils have the same number of mosquitoes due to
oviposition preference or due to egg and larval movement during rainfall.
The Frank et al. (1977) study determined that there was a correlation between the size of
the bromeliad and the number of mosquitoes within the bromeliad. The Frank et al. (1977)
experiment used the total water holding capacity of the bromeliad as the indicator of the size of
the bromeliad. In the current research, the actual amount of water within the bromeliad, not the
total holding capacity, was used to measure the densities of mosquitoes per milliliter of extracted
fluid. While the quantity of water actually present is probably not as good an indicator of size as
the total water holding capacity of the bromeliad, it gives an approximation of the habitat
available to the mosquito at the time of sampling.
There was positive correlation between the amount of water present in the bromeliad and
the number of mosquitoes extracted, although its strength varied between mosquito species.
Aedes albopictus had a stronger positive correlation with water volume than Wyeonzyia spp.,
which is possibly due to the fact that Wyeomyia spp. are bromeliad specialists, while A.
albopictus is a container generalist. Wyeomyia spp. are probably better adapted to the variable
bromeliad environment, and thus can thrive under a variety of bromeliad conditions, including
reduced water. For instance, Frank and Curtis (1977) hypothesized that the prolonged larval
development of W. vanduzeei under reduced food conditions was an adaptation to the variable
habitat in bromeliads. Aedes albopictus on the other hand, can also have prolonged larval
development with reduced resources (Barrera and Medialdea 1996), but probably cannot prolong
its life as long as Wyeomyia spp.. Thus, under variable conditions A. albopictus females have to
either be more selective for environments with more water for their oviposition, or risk a
reduction in survivorship due to desiccation or decreased space. Conversely, ovipositing
Wyeomyia spp. females do not have to be as selective for the amount of water in the bromeliad
habitat, as their larvae can tolerate alterations in water conditions.
In the habitat study Wyeomyia spp., but not A. albopictus, differed in mean densities under
different canopy types. It is possible that with more samples, A. albopictus would have also
differed in mean densities, because the total number ofA. albopictus recovered from all plants
was small. Since the palm trees offered less protection from the sun, the bromeliads under palms
were exposed to more insolation. In a study by Frank and O'Meara (1985), W. mitchellii showed
a preference for shaded habitats. In the current study, on two occasions the larvae collected from
most sites were identified to species. During both examinations about 80-85% of the Wyeomyia
spp. collected were W. mitchellii with the remaining being W. vanduzeei. Thus, the decrease in
density of Wyeomyia spp. in the palm habitat was most likely due to the preference of Wyeomyia
mitchellii for shaded areas.
Even though Wyeomyia spp. had the higher mean density in all plants, Wyeomyia spp. had
a higher mean density in B. pyramnidalis, while A. albopictus had a higher mean density in N.
spectabilis. In the following chapter, oviposition and larval competition experiments were
conducted to explain the causes of the differing relative abundances of Wyeomyia spp. and A.
albopictus in these two bromeliad species.
Table 5-1. Analysis of variance for the density of mosquitoes per plant based on month (month
of collection), site, and bromeliad species (N. spectabilis vs.B. pyramidalis).
df F-stati stic P-value
Month 6,472 0.27 0.53
Site 7,472 12.36 <0.01
Bromeliad species 1,472 7.92 <0.01
A repeated measures ANOVA was applied to the month variable.
Table 5-2. Analysis of variance for the density of mosquitoes per mL based on month (month of
collection), site, and bromeliad species (N. spectabilis vs. B. pyramidalis).
df F-stati stic P-value
A repeated measures ANOVA was applied to the month variable.
Table 5-3. Analysis of variance for the density of each mosquito species per mL of water in each
Mosquito Bromeliad Mean & SE (mos/mL) F- Stati sti c df P-value
A. albopictus N. spectabilis 0.0310.11 7.16 1,399 <0.01
B. pyramidalis 0.000310.0003
Wyeomyia spp. N. spectabilis 0.0910.01 42.53 1,399 <0.01
B. pyramidalis 0.2610.02
These means are representative of the data taken after February 2006. All samples
including those with no mosquitoes were included in this analysis.
Table 5-4. Mean densities of mosquitoes (mos/plant) in B.pyramidalis and N. spectabilis based on location within the plant and plant
Species Location albopictus Wy. spp. Total Total Wy spp. Total mosquito
Mean & SE
Mean & SE
Mean & SE
2.58 & 0.39
0.11 & 0.04
Mean & SE
Mean & SE
1.67 & 0.32
0.92 & 0.16
0.06 & 0.03
0.05 & 0.02
2.53 & 0.44
3.13 & 0.41
7.87 & 0.54
8.62 & 0.60
5.67 & 0.77
16.48 & 1.01
8.30 + 0.89
16.60 & 1.02
:. F1,472=42.53, P<0.01; 2: F1,472= 18.07, P<0.01; 3: F1,472= 18.51, P<0.01.
Table 5-5. Analysis of variance table for the effects of month and site on mos/mL and
Mosquito Effect F -stati sti c df P-value
mos/mL A. albopictus Month 9.17 4,378 <0.01
mos/mL Wyeomyia spp. Month 2.07 4,378 0.07
mos/plant A. albopictus Month 12.36 6,472 <0.01
mos/plant Wyeomyia spp. Month 8.74 6,472 <0.01
mos/plant A. albopictus Site 2.15 8,472 0.03
mos/plant Wyeomyia spp. Site 10.18 8,472 <0.01
Table 5-6. Analysis of variance results for the effect of month and sites on mosquito abundances
(mos/plant) by bromeliad species.
F-stati stic df P-value
Table 5-7. Average (+SE) amount of water extracted from each plant species.
N. spectabilis B. pyramnidalis N. spectabilis B. pyramnidalis
lateral axil lateral axil central axil central axil
109.12 & 5.80
79.09 & 2.90
Mean water 56.04 & 2.96 39.03 & 2.36 53.08 & 3.81
(mL) & SE
ANOVA results F1,198= 3.02; P<0.01- by plant species
33.28 & 1.45
F1,198= 47.59; P<0.01- by plant species
Table 5-8. A nested ANOVA table for effects of Oak vs. Palm canopy and bromeliad species (B.
pyramnidalis vs. N. spectabilis) on Wyeomyia spp. densities by mos/plant.
Oak vs. Palm
B. pyramnidalis vs. N. spectabilis
Oak vs. Palm
(B. pyramnidalis vs. N. spectabilis)
Table 5-9. Means and SE of densities ofA. albopictus and Wyeomyia spp. by location within
plants under oak and palm canopies.
Mean & SE
Wyeomyia spp. Oak B. pyramnidalis central axil 15.73 & 3.20
B. pyramnidalis lateral axil 7.80 & 1.16
N. spectabilis central axil 3.85 & 0.92
N. spectabilis lateral axil 4.08 & 0.73
Wyeomyia spp.Palm B. pyramnidalis central axil 1.00 + 0.50
B. pyramnidalis lateral axil 0.75 & 0.41
N. spectabilis central axil 0.57 & 0.43
N. spectabilis lateral axil 0. 14 & 0. 14
A. albopictus Oak B. pyramnidalis central axil 0
B. pyramnidalis lateral axil 0
N. spectabilis central axil 0
N. spectabilis lateral axil 0
A. albopictus Palm B. pyramnidalis central axil 0.13 & 0.13
B. pyramnidalis lateral axil 0.38 10.26
Figure 5-1. Map of collection sites. Figure adapted from O'Meara et al. 1995b, Fig. 1, Pg. 218.
Photo A Photo B
Figure 5-2. Location of water samples. The white arrow indicates the location where water samples were taken from the central axil
(Photo A) and the lateral axils (Photo B) of a N. spectabilis.
, 0 -\
-* eo l
-7 -Bi l
-6 Bi l
Fiue53 h otl endeste fWemi p.adA.abpcu nN pcaii
an .prmdlsa ahste ro asrpeetS f h en tec ie o
B.prmdlsn 4 lns(28 .5 lnssmldprmnha ahst
(ma S),adfrN pcaii =28pat 95 .0pat ape e
mot tec it ma E)
I I I I
-* A. albopictus
-o Wyeomyia spp.
b bd d
Figure 5-4: The monthly mean densities of Wyeone ia spp. and A. albopictus in N. spectabilis at
each site. Error bars represent SE of the means at each site. Means without a
common letter written above indicate significant differences for Wyeone ia spp..
Means without a common letter written below indicate significant differences for A.
SQ I b b
0- 10 _c
n a b b b b bb
-o Wyeomyia spp.
Figure 5-5. The monthly mean densities of Wyeongia spp. and A. albopictus in B. pyramnidalis at
each site. Error bars represent SE of the means at each site. Means without a
common letter written above indicate significant differences for Wyeongia spp..
Means without a common letter written below indicate significant differences for A.
OVIPOSITION AND LARVAL DEVELOPMENT OF Aedes albopictus IN TWO EXOTIC
SPECIES OF BROMELIAD IN SOUTHERN FLORIDA
Oviposition behavior by the female determines the aquatic habitat of the mosquito' s larval
stages (Clements 1999). Thus, it is not surprising that most ecological theories on oviposition
selection are generally based on the assumption that adults choose oviposition sites that
maximize the survivorship and reproductive output of their offspring and minimize their
development time (Srivastava and Lawton 1998). At least for some herbivorous insects, there is
some indication that fitness considerations of ovipositing females are also important in choice of
oviposition sites (Mayhew 2001). Whatever the ecological significance of oviposition
preference, numerous experiments have explored the cues that insects use to select oviposition
Culicids select oviposition sites based on visual, olfactory and tactile cues (Bentley and
Day 1989). The most common visual ovipositional stimuli of studied culicids are color cues.
Members of the genus Toxorhynchites are often attracted to black colored containers (Hilburn et
al. 1983, Jones and Schreiber 1994, Collins and Blackwell 2000), as are many species of the
genus Aedes (Beckel 1955, Wilton 1968). Culex mosquitoes are most often found to be attracted
to black and red (Dhileepan 1997), although color preference cannot be generalized by genus.
Aedes triseriatus are attracted to darker colors in the blue spectrum (Williams 1962, McDaniel et
al. 1976), and A. aegypti responded most strongly to black artificial bromeliads (Frank 1985).
There are also many chemical ovipositional cues originating from the habitat and from
conspecifics, predators, parasites and other species. There are numerous studies which indicate
that mosquitoes orient to cues from the organic material within the habitat (reviewed by
Clements 1999). For instance, in a laboratory study conducted by Wilton (1968), A. triseriatus
oviposited significantly more eggs in water collected from a treehole than in distilled water of
similar color. Culex spp. have also been found to respond to habitat specific chemical cues
derived from bacteria or the organic matter of the habitat (Ikeshoji et al. 1967, Beehler et al.
1994). Species of other mosquito genera also orient to chemical oviposition cues derived from
the habitat (Ikeshoji and Mulla 1970b, Bentley et al. 1979, Millar et al. 1992, Beehler et al.
1994). Habitat humidity (Kennedy 1942), salinity (Navarro et al. 2003), or algal content (Bond et
al. 2005) also influenced the oviposition site selection of mosquitoes.
Mosquitoes often respond to the presence of conspecifies and other species when choosing
oviposition sites. Some of the first olfactory studies conducted on Culicidae involved the
discovery of chemical oviposition cues in the egg rafts Culex spp. (Hazard et al. 1967, Ikeshoji et
al. 1967, Osgood 1971, Bruno and Laurence 1979, Laurence and Pickett 1982). In the Culex egg
raft experiments, female mosquitoes were attracted to oviposition sites that contained conspecific
eggs. Other mosquito species have shown a preference for ovipositing in the presence of
conspecific larvae (Soman and Reuben 1970, Bentley et al. 1976, Allan and Kline 1998, Mokany
and Shine 2003), or larvae of other species (Bentley et al. 1976, Allan and Kline 1998).
Mosquitoes are hypothesized to be attracted to oviposit in habitats with conspecifics, or other
species, because the presence of other mosquitoes is a possible indication of a successful larval
habitat (Clements 1999).
Some culicids have ovipositional preferences related to the presence or absence of
predators. In a study conducted by Munga et al. (2006), Anopheles gamnbiae avoided ovipositing
in containers that had been preconditioned with the waste of either of two predators. Avoidance
of oviposition in the presence of predators has also been seen for Ochlerotatus australis
(Mokany and Shine 2003), Culiseta longiareolata (Blaustein et al. 2004), and Aedes
taeniorhynchus (Ritchie and Laidlawbell 1994).
Once eggs are laid, the conditions within the habitat influence larval development. Since
A. albopictus is a container-inhabiting mosquito, variations in natural and artificial container
types occupied by this species influence larval development and survivorship. In treehole
communities, resource quantity within the habitat can influence larval development and
survivorship (Fish and Carpenter 1982, Leonard and Juliano 1995, Walker et. al 1997) and also
influence the community structure (Srivastava and Lawton 1998). The permanence of the habitat
also influences the development of larvae (Blaustein and Chase 2007), with drying of the habitat
often decreasing larval survivorship.
In southern Florida, A. albopictus is common in artificial or natural containers (O'Meara et
al. 1995a) and is occasionally found inhabiting bromeliads (O'Meara et al. 1995b). Within these
bromeliad habitats A. albopictus often co-occurs with other species such as W. mitchellii and W.
vanduzeei, and occasionally with A. aegypti, Aedes baha~nensis, Culex quinquefasciatus and
ulex biscaynesis (Frank 1985, O'Meara et al. 1995b, O'Meara et al. 2003). Past studies on
oviposition by A. albopictus have demonstrated a preference for darker colors (Gubler 1971, Yap
1975), and a lack of response to the presence of the bromeliad specialists W. vanduzeei and W.
naitchellii (Lounibos et al. 2003). Wyeonzyia vanduzeei showed a preference for oviposition into
flowering Tillandsia~~ll1~~~~111~~~ utriculata (Frank and O'Meara 1985).
During the field survey discussed in Chapter 5 of this thesis, there was a significant
difference between the densities of A. albopictus larvae in different species of bromeliads. Of
the two species of bromeliads that were the focus of the field survey, N. spectabilis had
significantly higher densities ofA. albopictus than B. pyra~nidalis. In order to understand the
differences in mosquito densities in these two plant species, an oviposition experiment was
implemented to determine whether oviposition was a factor that contributed to the higher
densities ofA. albopictus in N. spectabilis, and a larval development and survivorship
experiment was conducted to see whether the growth and survivorship ofA. albopictus differed
between N. spectabilis and B. pyramnidalis.
Materials and Methods
To determine whether A. albopictus preferentially oviposits into N. spectabilis in
comparison to B. pyramnidalis, an experiment was conducted in which 15 gravid A. albopictus
females were added by mouth aspirator to each of ten cages containing both one N. spectabilis
and one B. pyramnidalis bromeliad.
In accordance with approved animal-care protocol (VB-17 proj ect of the University of
Florida), each female had been bloodfed on a chicken five days before the start of the experiment
and held without an oviposition site. Only mosquitoes that appeared fully engorged were
removed from the main colony cage, and then transferred to a cage kept at 260C in a climate-
controlled insectary, with access to a 20% sucrose-water solution for the five days prior to the
start of the experiment. The colony from which the females were selected originated from
collections from southern Florida.
The N. spectabilis plants in the experiment were collected from one residence in the city
of Gotha in central Florida, and the B. pyramnidalis plants were collected from multiple
residences in Vero Beach and Fort Pierce, Florida. The N. spectablis plants chosen for the
experiment all contained partially submerged, nidulate inflorescences in the central tank
(Benzing 2000) and, thus, held larger amounts of water than non-flowering individuals. Because
N. spectabilis with inflorescences were the plants sampled in the field survey conducted in
Chapter 5, using flowering individuals for experiments may help to explain some of the findings
in that chapter. On the other hand, the B. pyra~nidalis in this experiment did not have
inflorescences, as their formation often destroys the water holding capacity of this species of
bromeliad. Non-flowering B. pyra~nidalis were also the only B. pyra~nidalis sampled in the field
surveys of Chapter 5.
Each plant was thoroughly washed by first rinsing in a bucket of water and then by
spraying with a high powered hose. Due to the fact that it is very difficult to remove all of the
leaves and organisms from the bromeliads, all plants were allowed to dry in an air-conditioned
laboratory from 8 to 10 days to ensure the death of all organisms due to desiccation. The plants
were then rewashed with a hose to remove any remains resulting from the drying.
Each replicate was enclosed in a 1.0 m wide and 0.76 m high pyramidal cage located
within a outdoor screened enclosure constructed for studies of mosquito flight behavior
(Bidlingmayer 1977). Within each cage, specimens of each bromeliad species were paired,
albeit the N. spectabilis were always slightly bigger. 1.0 g of chopped, dried (48 h at 750C) Q.
virginiana oak leaves, and 150 mL of sieved bromeliad water collected from both B. pyra~nidalis
and N. spectabilis bromeliads from the Vero Beach and Fort Pierce, Florida areas were added to
the central tank of each bromeliad. The water of each species was pooled, so the water added to
each plant was a mixture from both bromeliad species. The water was sieved through a 130 Clm
mesh to remove macroinvertebrates and detritus. The number of water holding leaves was
counted as an estimate of plant water holding capacity.
The experiment was run for 7 days, after which the remaining water in each bromeliad
was removed by a pipette, and each leaf containing water was rinsed with tap water into a metal
rearing pan. Each pan was kept for 10 days at 260C in a climate-controlled rearing room and
covered with a sheet of glass to prevent possible oviposition from stray mosquitoes in the rearing
room. On each day, the pans were checked for newly hatched A. albopictus, and the sums of
hatched larvae in the two bromeliad species were regarded as products of recent oviposition and
compared by a paired t-test.
Larval Development and Survivorship Experiment
Fourteen plants, 7 B. pyra~nidalis and 7 N. spectabilis, from the previous experiment
were washed in the manner as described above, and 1.0 g of chopped oak leaves (Q. virginiana),
and 150 mL of sieved bromeliad water were added to the central tank of each bromeliad. Then
after 20 4th instar Wyeonzyia spp. and 20 1st instar A. albopictus had been added into the center
tank, the plants were placed into metal cages in a 260C climate controlled rearing room. After 3
days, 50 mL of water was added to the plants to compensate for evaporation. After 6 days, all of
the contents of the bromeliads were rinsed into a pan in the manner as the oviposition
experiment. Variations between plant species in the number and developmental stage of
surviving A. albopictus were analyzed by ANOVA in SAS (2002). The survivorship data was
arcsine square root transformed to meet the homogeneity of variances assumption of ANOVA.
Aedes albopictus oviposited significantly more (t9= 2.95, P<0.01) eggs into N.
spectabilis (mean & SE=33.5016.38) than B. pyra~nidalis (mean & SE=15.80 & 5.29).
Neoregelia spectabilis had significantly more (t9= 3.33, P<0.01) (mean & SE= 9.60 + 0.70)
water-holding axils than B. pyra~nidalis (mean & SE= 7.1 & 0.46).
Larval Development and Survivorship Experiment
A. albopictus had a significantly higher (F1,12= 67.41, P<0.01) average instar number
within N. spectabilis (mean & SE=3.96 & 0.22) than in B. pyra~nidalis (mean &
SE=2.3110.08).The mean survivorships of A. albopictus in N. spectabilis (0.56 & 0.08 (SE) )and
B. pyra~nidalis (0.67 & 0.04 (SE)) were not significantly different (F1,12= 1.43, P=0.25).
While there was significantly more oviposition by A. albopictus in N. spectabilis, this
experiment did not consider the cause of this ovipositional preference The difference could be
based on coloration differences between N. spectabilis and B. pyra~nidalis. N. spectablis is dark
green with dark purple tips as compared to the entirely light-green B. pyra~nidalis. In an
experiment conducted by Frank (1985), other species of mosquitoes occasionally found in
bromeliads, C. quinquefasciatus, A. aegypti, and Toxorhynchites rutilus, preferentially oviposited
in darkly colored artificial bromeliads, as opposed to lighter green artificial bromeliads. In this
same study, Wyeonzyia spp. preferred to oviposit in the lighter green artificial bromeliads.
Therefore, it seems possible that A. albopictus, like other container-generalist mosquito species,
has a preference for the darker colored bromeliad N. spectabilis.
Another visual cue besides color that may have affected the experiment was the size of the
plants. Generally in the field, and in this experiment as well, N. spectabilis is a horizontally
larger plant with much longer leaves. B. pyra~nidalis tends to be smaller overall with shorter
leaves and a more vertically dominant stature than horizontally dominant. The paired t-test
demonstrated that the N. spectablis also has more water holding axils per plant than the B.
pyra~nidalis plants in this experiment.
Bentley and Day (1989) stated that specialist mosquitoes, those which have more
restricted ovipositional sites, such as crab holes, bromeliads and natural or artificial containers,
tend to rely strongly on visual cues to aid in the identification of ovipositional sites. While this
statement may also favor the color difference hypothesis for the observed ovipositional
preference of A. albopictus for N. spectablis, the fact that there was a difference in the size of the
two plants could also be a factor. Neoregelia spectabilis, the larger of the two bromeliad species,
may be visually more attractive to A. albopictus, regardless of color. This larger size is not an
artifact of the experiment, because in nature N. spectabilis are commonly larger than B.
pyra~nidalis. Therefore, although the difference may be due to size, there are still important
ecological implications to this difference.
On the other hand, visual cues may be of little importance to the ovipositional decision of
A. albopictus in bromeliads. Aedes albopictus could be responding to habitat-based chemical
cues unique to each plant species. For instance, the submerged nidulate inflorescence ofN.
spectaibilis may provide additional nutrients or chemical compounds to the central tank water,
while B. pyra~nidalis does not receive this input, because it creates tall, stalky inflorescenses.
Since the water from both species was pooled, each species received water originating from both
B. pyra~nidalis and N. spectabilis. Therefore, it is unlikely that chemical compounds within the
water contributed to the oviposition preference ofA. albopictus for N. spectabilis in this
experiment. The experiment ran for 7 days, so only chemical cues that accumulated in this short
period of time could have affected the preference of A. albopictus for N. spectabilis.
The larval competition experiment within N. spectabilis and B. pyra~nidalis showed that
A. albopictus can develop at different rates based on the species of bromeliad. The multitude of
structural differences in size and shape between the two plant species could contribute to the
developmental differences of A. albopictus within the two plants. In Chapter 2 of this thesis, and
in treehole microcosms study by Broadie and Bradshaw 1991, surface area to volume ratio was
not a factor influencing the development rate ofAedes mosquitoes. So, it is possible that other
structural differences account for the increased development rate ofA. albopictus in N.
Neoregelia spectabilis has a sunken inflorescence which was decaying during this
experiment, while B. pyramnidalis did not have this decaying inflorescence. Therefore, it is
possible that this sunken inflorescence provided additional nutrients that increased the food
resources available to A. albopictus. Further experiments are necessary to elucidate the causes of
the developmental differences ofA. albopictus seen in these two bromeliad species.
This experiment shows the importance of bromeliad species in the distribution of A.
albopictus. So far in North America, A. albopictus has been found to be infected with LaCrosse
(Gerhardt et al. 2001), Eastern Equine Encephalitis (EEE) (Mitchell et al. 1992), and dengue
viruses (Ibafiez Bernal et al. 1997), and transmits Dengue (DEN) in other areas of the world
(Hawley 1988). As this mosquito species is a health concern, then the understanding of its
ovipositional preferences within bromeliads can be beneficial to determining the risk of
transmission of these viruses to human populations. In Brazil, bromeliad eradication projects
have been implemented to eliminate these supposed development sites of dengue vectors
(Benzing 2000). With the higher populations of A. albopictus in northern Florida where
Wyeomyia spp. is absent (O'Meara et al. 1995b), further studies on the ovipositional preferences
of A. albopictus in bromeliad species could help reduce risk by applying control measures only
to bromeliad species that are of highest concern.
In aquatic container habitats, high densities of mosquito larvae can increase development
time and reduce survivorship and pupal weight (e.g. Frank and Curtis 1977, Mogi 1984, Livdahl
1982, Broadie and Bradshaw 1991). In the surface area to volume ratio experiment of Chapter
2, and a larval competition experiment conducted by Lounibos et al. (2003), A. albopictus
suffered reduced growth and survivorship in the presence of larger 4th instar Wyeomyia spp..
Resource competition is common among mosquito larvae, and interference competition in
the forms of chemical interference (Dye 1984, Sunhara and Mogi 2002, Bedhomme et al. 2005),
encounter competition (Dye 1984, Roberts 1998, Broadie and Bradshaw 1991), and facultative
or intraguild predation (e.g. Reisen and Emory 1976, Koenekoop and Livdahl 1986, Koenraadt et
al. 2003) also influence larval mosquito communities. The results of this study suggest that
encounter competition, not chemical interference or predation, causes the reduced growth and
survivorship of early instar A. albopictus in the presence of 4th instar Wyeomyia spp. in
bromeliads. Increasing the habitat complexity in an artificial bromeliad increased the
developmental rate ofA. albopictus in the presence of Wyeomyia spp.. However, because the
effect of habitat complexity on the development time ofA. albopictus in the absence of
Wyeomyia spp. was not explored, the relationship of habitat complexity to A. albopictus
development may be unrelated to interactions with Wyeomyia spp.
Even though large numbers of mosquitoes were used in the chemical interference
experiments, excretory products of Wyeomyia spp. did not affect the growth or survivorship of
A. albopictus in bromeliads. Neither the pH nor ammonia concentration of bromeliad water
differed before and after the experiment. Because bromeliads assimilate ammonia from their