PLACE IN RETURN Box to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 KIProj/Aoc8-PralelRC/Dateouejndd

 

 

MOSQUITO PRODUCTION AND MICROBIAL DIVERSITY IN CONTAINER
HABITATS

By

Kirsten Suzanne Pelz

A DISSERTATION
Submitted to
Michigan State University
in partial fulfilhnent of the requirements
for the degree of
DOCTOR OF PHILOSOPHY
Department of Entomology and Program in Ecology, Evolutionary Biology and Behavior

2008

ABSTRACT

MOSQUITO PRODUCTION AND MICROBIAL DIVERSITY IN CONTAINER
HABITATS

By

Kirsten Suzanne Pelz

Container-breeding mosquitoes comprise approximately 40% of known mosquito
species. In addition to man-made containers, including tires and cemetery vases, many of
these mosquitoes reside in natural container habitats, such as water-filled tree holes.
Detritus is a key component of larval nutrition and its availability and quality directly
relate to the production of adult mosquitoes. Microbial metabolism incorporates nutrients
from detritus, which mosquitoes then procure via direct consumption of microorganism
in biofilms and in the water column. The successful emergence of adults depends on the
consumption of these microbial communities; therefore, I have examined the interaction
of several container dwelling mosquito species, Ochlerotatus triseriatus, Aedes
albopictus and Aedes aegypti in order to evaluate the contributions of microbial
community dynamics to mosquito development. The studies in this dissertation were
. designed to integrate microbial community level dynamics with broader ecological
processes associated with tree hole communities, including decomposition, competition,
and facilitation. Using terminal restriction fragment polymorphism (T-RF LP) analysis
and sequencing, I describe herein changes in the structure of bacterial and fungal
communities in response to container type, mosquito density, and macroinvertebrate

community composition.

ACKNOWLEDGMENTS

My journey towards a doctoral degree has been an experience that although
frustrating or difficult at times, has served to enhance my sense of wonder about the
natural world. I have been fortunate in that my research program has required that I
acquire a more than passing familiarity with a variety of biological disciplines that are
often separated under the traditional structure imposed by not only a university
departmental structure, but the field of science as a whole. I have had an uncommonly
integrated experience that has strengthened my desire to be a lifetime student of biology.

The pursuit of this degree has been a personal challenge that would not have been
possible without the continuous support of my parents, who have provided love and
unconditional support throughout my life. I have been most fortunate as well in having
the encouragement, love, and support of my husband, Lukasz. I do not know how others
survive a doctoral program, but I am certain that I could not have done so without having
him in my life as both a wise colleague and a best fi'iend.

Many people have provided assistance in the laboratory, for which I am truly
grateful: Blair Bullard, Bill Morgan, Renee Bloome, Bob Burns, Kevin Vogel, and Ryan
Heinke. There is one person, however, who stands out among the wonderful people I
have worked with over the last three years. I had the pleasure of working with Lane
LaFleur Frasier for the majority of my Ph.D. Uncommonly wise beyond her years and
skilled in the lab, Lane has been both an exceptional employee and a good friend.

Finally, I am very appreciative of the guidance provided by my major professors,
Mike Kaufman and Ned Walker. I have enjoyed the wit and intelligence they have

contributed to our interactions throughout my program.

iii

TABLE OF CONTENTS

LIST OF TABLES ............................................................................................................ vii

LIST OF FIGURES ......................................................................................................... viii

CHAPTER 1

INTRODUCTION ............................................................................................................... 1
Systematics .............................................................................................................. 2
Medical Significance ............................................................................................... 3
Distribution and Life History ................................................................................... 4
Larval habitat and Nutrition ..................................................................................... 5
Ecological Factors Effecting Adult Mosquito Production ....................................... 7
Research Objectives and Rationale .......................................................................... 8

CHAPTER 2

EFFECTS OF LEAF QUALITY AND LABILE COMPONENTS ON GROWTH
PARAMETERS OF OCHLEROTA TUS TRISERIATUS (SAY) (DIPTERA:

CULICIDAE) AND ASSOCIATED MICROBIOTA ....................................................... 10
Introduction ............................................................................................................ 1 1
Materials and Methods ........................................................................................... 15

Microcosm construction ............................................................................. 15
Experiment 1 .............................................................................................. 15
Experiment 2 .............................................................................................. 16
Experiment 3 .............................................................................................. 17
Chemical analyses ...................................................................................... l7
Bacterial abundance ................................................................................... l 8
Bacterial productivity ................................................................................. 1 8
Statistical analysis ...................................................................................... 19
Results .................................................................................................................... 20
Experiment 1 .............................................................................................. 20
Experiment 2 .............................................................................................. 26
Experiment 3 .............................................................................................. 37
Discussion .............................................................................................................. 39

CHAPTER 3

HABITAT-SPECIFIC CHANGES IN MICROBIAL COMMUNITY DIVERSITY

ASSOCIATED WITH CONTAINER-BREEDING MOSQUITOES ............................... 42
Introduction ............................................................................................................ 43
Materials and Methods ........................................................................................... 44

Field ........................................................................................................... 44
Laboratory .................................................................................................. 46
Molecular Analysis .................................................................................... 47
Bacterial and Fungal rRNA gene sequences .............................................. 48

iv

T-RFLP Analysis ....................................................................................... 49

Statistical Analyses .................................................................................... 49
Results .................................................................................................................... 50
Tree holes ................................................................................................... 50
Tree hole microcosms ................................................................................ 57
Tires ........................................................................................................... 62
Tire Microcosms ........................................................................................ 69
Discussion .............................................................................................................. 73
CHAPTER 4
MULTITROPHIC INTERACTIONS IN TREE HOLE COMMUNITIES ARE
FACILITATED SCIRTID BEETLES ............................................................................... 78
Introduction ............................................................................................................ 79
Materials and Methods ........................................................................................... 82
Experimental Design .................................................................................. 82
Sampling .................................................................................................... 83
Bacterial abundance and productivity ........................................................ 83
Fungal biomass .......................................................................................... 84
Enzyme activity ......................................................................................... 85
Microbial community analysis ................................................................... 85
Statistical analysis ...................................................................................... 86
Results .................................................................................................................... 87
T-RFLPS ..................................................................................................... 99
Discussion ............................................................................................................ 102
CHAPTER 5

COMPETITIVE INTERACTIONS BETWEEN OCHLEROTA TUS TRISERIA TUS
(SAY) AND AEDES ALBOPICTUS (SKUSE) (DIPTERA: CULICIDAE) ARE
INFLUENCED BY HABITAT-ASSOCIATED MICROBIAL COMMUNITY

DYNAMICS .................................................................................................................... 109
Introduction .......................................................................................................... 1 10
Materials and Methods ......................................................................................... 114

Microcosms .............................................................................................. 1 14
Samples .................................................................................................... 1 15
Mosquito measurements .......................................................................... 1 15
Bacterial Abundance ................................................................................ 1 15
Bacterial Productivity .............................................................................. l 16
T-RFLP Analysis ..................................................................................... 1 16
Statistical Analysis ................................................................................... 1 18
Results .................................................................................................................. 1 18
Mosquito Production ................................................................................ 118
Bacterial Abundance and Productivity .................................................... 119
T-RFLPS ................................................................................................... 128
Discussion ............................................................................................................ 135
CHAPTER 6

INTERSPECIFIC MICROBIAL RESOURCE UTILIZATION IN AEDES ALBOPICTUS

AND AEDES AE GYPT I ................................................................................................... 141
Introduction .......................................................................................................... 142
Materials and Methods ......................................................................................... 145

Microcosms .............................................................................................. 145
Sampling .................................................................................................. 146
Bacterial Production ................................................................................. 146
Microbial community analysis ................................................................. 147
Statistical Analysis ................................................................................... 148
Results .................................................................................................................. 148
Mosquito Production ................................................................................ 148
Bacterial Productivity .............................................................................. 153
T—RFLPs ................................................................................................... l 56
Discussion ............................................................................................................ 162

CONCLUSIONS .............................................................................................................. 168

APPENDIX 1.. ................................................................................................................ 171
Appendix 1.1. Record of Deposition Of Voucher Specimens .............................. 172
Appendix 1.2. Voucher Specimen Data ............................................................... 173

APPENDIX 2. SUMMARY OF AN OVA RESULTS FOR PRINCIPAL COMPONENT
ANALYSIS OF MICROBIAL COMMUNITIES ASSOCIATED WITH MICROCOSMS
IN CHAPTER 3.. ............................................................................................................ 174

LITERATURE CITED .................................................................................................... 184

vi

LIST OF TABLES

Table 2.1. Multivariate (MAN OVA) and univariate (ANOVA) hypothesis test results for
all factors and their interactions for mosquito production variables in experiment 1.
ANOVA results for each variable are shown for factors with significant MANOVA
(P<0.05) ............................................................................................................................. 22

Table 2.2. Multivariate (MANOVA) and univariate (ANOVA) hypothesis test results for
all factors and their interactions for mosquito production variables in experiment 2.
AN OVA results for each variable are shown for factors with significant MAN OVA
(P<0.05) ............................................................................................................................. 26

Table 2.3. Experiment 2 multivariate (MANOVA) and univariate (AN OVA) hypothesis
test results for all factors and their interactions for productivity and abundance of bacteria

on leaf surfaces and in the water column. ANOVAs were performed for significant
ANOVA factors (P<0.05) .................................................................................................. 27

Table 2.4. Multivariate (MAN OVA) and univariate (AN OVA) hypothesis test results for
all factors and their interactions for nitrogen and phosphorus concentration in experiment
2. AN OVA’s were performed for significant ANOVA factors (P<0.05) ......................... 34

Table 2.5. MANOVA results for all factors and their interactions for mosquito
production variables in experiment 3. AN OVA results for each variable are shown for
factors with significant MANOVA (P<0.05). ................................................................... 37

Table 3.1. Analysis of variance results (ANOVA) for principle component (PC) values
obtained from relative abundance of T-RFLP fragments in bacterial (16S rRNA) and
fungal (188 rRNA) communities in tree holes. Shown are F values (F), degrees of
freedom ((11) and p values (P) for the main effect of sampling time for each substrate ....51

Table 3.2. Analysis of variance results (AN OVA) for principle component (PC) values
obtained from relative abundance of 165 rRNA gene T-RFLP fragments in tree hole
microcosm communities and showing F values (F), degrees of freedom ((10 and p values
(P) for the main effects of sampling time and mosquito density on each substrate .......... 57

Table 3.3. Analysis of variance results (ANOVA) for principle component (PC) values
obtained from relative abundance of T-RF LP fragments in bacterial (16S rRNA)
communities in tires. Shown are F values (F), degrees of freedom (di) and p values (P)
for the main effect of sampling time and mosquito density for each substrate ................. 63

Table 3.4. Analysis of variance results (ANOVA) for principle component (PC) values
obtained from relative abundance of T-RFLP fi'agments in fungal (18S rRNA)
communities in tires. Shown are F values (F), degrees of freedom ((11) and p values (P)
for the main effect of sampling time and mosquito density for each substrate ................. 64

vii

Table 3.5. Analysis of variance results (ANOVA) for principle component (PC) values
obtained from relative abundance of 16s rRNA gene T-RFLP fragments in tire
microcosm communities and showing F values (F), degrees of freedom (df) and p values
(P) for the main effects of sampling time and mosquito density on each substrate .......... 70

Table 4.1. Summary of AN OVA results for bacterial abundance and bacterial
productivity values from microcosm water column and leaf material .............................. 93

Table 4.2. Summary of ANOVA results for ergosterol and fungal degradation enzyme
concentrations in microcosm leaf material ........................................................................ 96

Table 5.1. Multivariate analysis of variance (MANOVA) results for emergence and mass
of A. albopictus and 00. triseriatus in microcosms. Total adults and females were
analyzed as separate dependent variables . ...................................................................... 121

Table 5.2. Multivariate analysis of variance (MANOVA) results for bacterial abundance
in microcosms on days 15 and 30 . .................................................................................. 125

Table 5.3. Multivariate analysis of variance (MANOVA) results for bacterial
productivity in microcosms on days 15, 30 and 60 . ....................................................... 125

Table 5.4. Analysis of variance (ANOVA) results for PCI values obtained from PCA

analysis of bacterial (16S rDNA) T-RF LP peak area data . ............................................ 128
Table 5.5. Analysis of variance (ANOVA) results for PCI values obtained from PCA

analysis of fimgal (lSS rDNA) T-RF LP peak area data . ................................................ 131
Table 5.6. Analysis of variance (ANOVA) results for PC2 values obtained from PCA

analysis of fungal (1 SS rDNA) T-RFLP peak area data . ................................................ 132
Table 5.7. Bacterial (16S) T-RF LP fragments with high factor loadings . ..................... 133
Table 5.8. Fungal (1 SS) T-RF LP fragments with high factor loadings .......................... 133
Table 6.]. ANOVA results for mosquito production variables ...................................... 150
Table 6.2. AN OVA results for bacterial productivity . ................................................... 154

Table 6.3. Summary of ANOVA results for principal component (PC) scores from leaf
surface bacterial (16S rRNA gene) T-RFLP profiles . .................................................... 157

Table 6.4. Summary of ANOVA results for principal component (PC) scores from leaf
surface fungal (lSS rRNA gene) T-RFLP profiles .......................................................... 158

viii

Appendix 2, Table 1. Summary of ANOVA results for PCI scores from water column
16S rDNA T-RFs digests (Block I) ................................................................................ 175

Appendix 2, Table 2. Summary of ANOVA results for PCI scores fi'om water column
16S rDNA t-RFs digests (Block 11). ............................................................................... 176

Appendix 2, Table 3. Summary of ANOVA results for PCI and PC2 scores from leaf
surface 16S rDNA t-RFs digests (Block I) ..................................................................... 177

Appendix 2, Table 4. Summary of ANOVA results for PCl and PC2 scores from leaf
surface 168 rDNA t-RFs digests (Block 11). .................................................................... 178

Appendix 2, Table 5. Summary of ANOVA results for PCI and PC2 scores from water
column 188 rDNA t-RFs digests (Block I). ..................................................................... 179

Appendix 2, Table 6. Summary of ANOVA results for PCI and PC2 scores fiom water
column 188 rDNA t-RFs digests (Block II) .................................................................... 180

Appendix 2, Table 7. Summary of ANOVA results for PCI and PC2 scores from leaf
surface 188 rDNA t-RFs digests (Block I) ...................................................................... 181

Appendix 2, Table 8. Summary of AN OVA results for PCI and PC2 scores from leaf
surface 18S rDNA t-RFs digests (Block II) ..................................................................... 182

ix

LIST OF FIGURES

Figure 2.1. Mosquito production variables from experiment 1. (A) Survival (20 initial
larvae). (B) Average female development time. (C) Average male development time. (D)
Average female weight. (E) Average male weight. Values are means :1: SE (n = 6 for all
variables in A, C, and E. n = 2-6 for variables in B and D) ............................................... 23

Figure 2.2. Leaf mass lost in experiment 1. Values are means t SE (n = 6) .................... 24

Figure 2.3. Mosquito production variables from experiment 2. (A) Survival (40 initial
larvae). (B) Average female development time. (C) Average male development time. (D)
Average female weight. (E) Average male weight. Values are means 1: SE (n = 6 for all
variables in A, C, and E. n = 1-6 for variables in B and D) . ............................................. 28

Figure 2.4. Water column bacterial productivity (leucine incorporation rate) response to
leachate, filtration, and leaf condition in experiment 2. (A) day 0 and (B) day 70. Values
are means :1: SE (n = 6). Values are means :1: SE (n = 6) .................................................... 30

Figure 2.5. Bacterial abundance (direct microscopic counts) response to leachate,
filtration, and leaf condition in experiment 2. Water column: (A) day 0 and (B) day 70.
Leaf surface: (C) day 0 and (D) day 70. Values are means d: SE (n = 6) .......................... 31

Figure 2.6. Water column chemistry for experiment 2 at days 0 (A,B), 12 (CD), and 70
(E,F). Shown are total nitrogen (A,C,E) and total phosphorous (B,D,F) .......................... 36

Figure 2.7. Mosquito production variables from experiment 3. (A) Average leaf mass
remaining. (B) Survival (40 initial larvae). (C) Average female development time. (D)
Average male development time. (E) Average female weight. (F) Average male weight.
Values are means :t SE (n = 6 for all variables in A, B, E, and F. n = 0-6 for variables in
C and D). Means within each column followed by the same letter are not significantly
different, (P > 0.05, Fisher’s Protected LSD Test) ........................................................... 37

Figure 3.1. Principal component analysis of bacterial 16S rRNA gene communities from
tree hole habitats. Panel A: Leaf, B: Water, C: Tile. PC axes 1 and 2 explained 24, 17,
and 26% of the variation for the respective substrates ..................................................... 52

Figure 3.2. Percentage of bacterial 16S rRNA gene sequences in the class taxonomic
level for composite samples taken from substrates in tree holes. A Leaf, B Water, C Tile
(n = 4 per composite) ........................................................................................................ 53

Figure 3.3. Principal component analysis of fungal ISS rRNA gene communities from
tree hole habitats. Panel A: Leaf, B: Water, C: Tile. PC axes 1 and 2 explained 20, 19,
and 28% of the variation for the respective substrates ..................................................... 54

Figure 3.4. Percentage of fungal ISS rRNA gene sequences in the class taxonomic level
(or above) for composite samples taken from substrates in tree hole habitats on day 15. A
Leaf, B Water, C Tile (n = 4 per composite). Inset legend refers to mosquito density ..... 55

Figure 3.5. Principal component analysis of bacterial l6S rRNA gene communities fiom
tree hole microcosms. Panel A: Leaf, B: Water, C: Tile. PC axes 1 and 2 explained 45,
50, and 69% of the variation for the respective substrates ................................................ 58

Figure 3.6. Percentage of bacterial 16S rRNA gene sequences in the class taxonomic
level for composite samples taken from substrates in microcosms simulating tree holes. A
Leaf, B Water, C Tile (n = 6 per composite) ..................................................................... 59

Figure 3.7. Percentage of fungal lSS rRNA gene sequences in the class taxonomic level

(or above) for composite samples taken from substrates microcosms simulating tree holes
on days 0, 15, and 30. AB Leaf, C,D Water, E,F Tile (n = 6 per composite). Panels A, C,
and E are microcosms with 0 larvae, B, D, and F are microcosms with 40 larvae ........... 60

Figure 3.8. Principal component analysis of bacterial 16S rRNA gene communities from
tire habitats. Panel A: Leaf, B: Water, C: Tile. PC axes l and 2 explained 20, 19, and
28% of the variation for the respective substrates . ........................................................... 65

Figure 3.9. Percentage of bacterial 16S rRNA gene sequences in the class taxonomic
level for composite samples taken from substrates in tire habitats. A Leaf, B Water, C
Tile (n = 4 per composite) ................................................................................................. 66

Figure 3.10. Principal component analysis of fungal lSS rRNA gene communities from
tire habitats. Panel A: Leaf, B: Water, C: Tile. PC axes 1 and 2 explained 24, 23, and
14% of the variation for the respective substrates ............................................................. 67

Figure 3.11. Percentage of fungal 18S rRNA gene sequences in the class taxonomic level
(or above) for composite samples taken from substrates in tire habitats on day 15. A Leaf,
B Water, C Tile (n = 4 per composite). Inset legend refers to mosquito density .............. 68

Figure 3.12. Principal component analysis of bacterial 16S rRNA gene communities
from tire microcosms. Panel A: Leaf, B: Water, C: Tile. PC axes l and 2 explained 29,
42, and 58% of the variation for the respective substrates ................................................ 71

Figure 3.13. Percentage of bacterial 16S rRNA gene sequences in the class taxonomic
level for composite samples taken from substrates in microcosms simulating tire habitats.
A Leaf, B Water, C Tile (n = 6 per composite) ................................................................. 72

Figure 4.1. Mosquito production variables from microcosms containing 40 0. triseriatus
larvae. . (A) Number of adults (B) Total adult mass. (C). Survival (40 initial larvae).

open characters represent microcosms with no scirtids filled characters represent 10
scirtids. Values are means i SE; n = 12 all variables . ...................................................... 89

xi

Figure 4.2. Proportion of leaf mass remaining in microcosms on day 50. S: Scirtid, M:
Mosquito, white bars: 0 M, grey bars : 10 M. Values are means :1: SE (N=24) ................. 90

Figure 4.3. Bacterial production and abundance. S: Scirtid, M: Mosquito, white bars: 0 S,
grey bars : 10 S. Values are means :t SE (N=24) .............................................................. 95

Figure 4.4. Fungal enzyme activity . S: Scirtid, M: Mosquito, white bars: 0 S, grey bars :
10 S. Values are means :1: SE (N=24) . .............................................................................. 97

Figure 4.5. Fungal biomass and enzyme activity . S: Scirtid, M: Mosquito, white bars: 0
S, grey bars : 10 S. Values are means t SE (N=24) . ........................................................ 98

Figure 4.6. PCA ordination of bacteria peak area data for leaf (A and B) and water (C
and D) samples taken from Block I (A and C) and Block 11 (B and D) microcosms on day
20. The amount of variation explained by each PC is indicted on the axes for each panel.
Treatments are represented by the following symbols: open, 0 scirtids; filled, 10 scirtids;
squares, 0 mosquitoes; circles, 40 mosquitoes ................................................................. 101

Figure 4.7. PCA ordination of fungal peak area data for leaf (A and B) and water (C and
D) samples taken from Block I (A and C) and Block 11 (B and D) microcosms on day 20.
The amount of variation explained by each PC is indicted on the axes for each panel.
Treatments are represented by the following symbols: open, 0 scirtids; filled, 10 scirtids;
squares, 0 mosquitoes; circles, 40 mosquitoes ................................................................ 102

Figure 5.1. Development time for two densities of A. albopictus and 0c. triseriatus in
microcosms. Low mosquito densities are represented by filled characters and high
densities by open characters. Panels A and B are females, C and D are males. Values are
means i SE, n = 3-4 for each point ................................................................................. 122

Figure 5.2. Number of adult A. albopictus and 0c. triseriatus emerged from microcosms.
Low mosquito densities are represented by filled characters and high densities by open
characters. Panels A and B are females, C and D are total number of adults. Values are
means :t SE, n = 3-4 for each point ................................................................................. 123

Figure 5.3. Mass of adult A. albopictus and 0c. triseriatus from microcosms. Low
mosquito densities are represented by filled characters and high densities by open
characters. Panels A and B are females, C and D are males. Values are means A: SE, n =
3-4 for each point . ........................................................................................................... 124

Figure 5.4. Bacterial abundance (direct microscopic counts) in microcosms containing
two densities of A. albopictus (A) and 0c. triseriatus (T), alone or under interspecific
competition. Low mosquito densities are represented by filled characters and high
densities by open characters. Panels A and C are water column samples, B and D are leaf
surface samples. Samples were taken on day 15 (A and B) and Day 30 (C and D). Values
are means iSE, n = 3-4 for each point ............................................................................. 126

xii

Figure 5.5. Bacterial productivity (leucine incorporation rates) in microcosms containing
two densities of A. albopictus (A) and 0c. triseriatus (T), alone or under interspecific
competition. Low mosquito densities are represented by filled characters and high
densities by open characters. Panels A and B reflect samples taken from the water
column, C and D are from leaf surfaces. Samples, taken on day 15 (A and C) and day 30
(B and D), are represented as means i SE, n = 3-4 for each point . ................................ 127

Figure 5.6. Principle component analysis ordination of bacterial T-RF LP peak area data
for microcosms containing two densities of A. albopictus (squares) and 0c. triseriatus
(circles), alone or under interspecific competition (triangles). High mosquito densities are
represented by filled characters and low densities by open characters. Panels A and B are
water column samples, C and D are leaf surface samples. Samples were taken on day 15
(A and C) and Day 30 (B and D). Values are means iSE, n = 3-4 for each point .......... 130

Figure 5.7. Principle component analysis ordination of fungal T-RFLP peak area data
for microcosms containing two densities of A. albopictus (squares) and 0c. triseriatus
(circles), alone or under interspecific competition (triangles). High mosquito densities are
represented by filled characters and low densities by open characters. Panels A and B are
water column samples, C and D are leaf surface samples. Samples were taken on day 15
(A and C) and Day 30 (B and D). Values are means iSE, n = 3-4 for each point .......... 134

Figure 6.1. Mosquito production variables for A. albopictus males (A and C) and females
(B and D). Average mosquito weight (A and B) and development time (C and D) are
shown as means :t SE. n = 6 for variables in A and B, n = 2-6 for variables in C and D .151

Figure 6.2. Mosquito production variables for A. aegypti males (A and C) and females (B
and D). Average mosquito weight (A and B) and development time (C and D) are shown
as means :1: SE. n = 6 for variables in A and B, n = 0-6 for variables in C and D.
Treatments means with different letters are significantly different following Bonferroni
correction for experiment wide error ............................................................................... 152

Figure 6.3. Bacterial productivity in microcosms on day 7 and day 14 following larval
addition. (A) Leaf surface. (B) Water column. Values are means i SE (n=5-6) ............. 155

Figure 6.4. PCA of TRFLP fragment peak areas from leaf surface bacterial (l6S rRNA
gene) communities. Total density of mosquitoes (0, 30, or 60) are compared for each
sampling date. The two component axes explain 48.1% of the variation ........................ 159

Figure 6.5. PCA of TRFLP fragment peak areas from water column bacterial (16S rRNA
gene) communities. Individual A. albopictuszA. aegypti treatments are compared. The two
component axes explain 60.6% of the variation .............................................................. 160

Figure 6.6. PCA of T-RFLP fragment peak areas from leaf surface fungal (1 SS rRNA

gene) communities. Comparisons of individual A. albopictus: A. aegypti treatments are
shown. The two component axes explain 27.9% of the variation ................................... 161

xiii

CHAPTER 1

INTRODUCTION

Container-dwelling forms represent approximately 40% of mosquito species, many of
which are important arbovirus vectors, including West Nile Virus, yellow fever, and
LaCrosse encephalitis (Laird 1988). Water-filled containers may be man-made (e. g. tires,
cemetery vases, flower pots) or plant-based. The latter, called phytotehnata (phyto =
plant, telmata = container), commonly occur as tree holes found either along tree trunks
or at the tree base formed by root outcroppings. Tree holes are an excellent subject for
studies of trophic interactions, due to the self-contained nature of the communities
residing therein. Many insects with aquatic larval stages make their home in tree holes.
Although typically dominated by mosquito larvae, non-mosquito representatives like
Helodes pulchella, Prionocyphon discoideus (Coleoptera: Scirtidae), and Culicoides
guttipemzis (Diptera: Ceratapogonidae) are also common (Barrara 1988, Paradise 2000).
The mosquito species in tree holes varies geographically. In Michigan, the primary
mosquito species residing in tree holes is the Eastern treehole mosquito, Ochlerotatus
triseriatus (Say). Recently, 0c. japom'cus has begun to invade Michigan tree holes
following its introduction into the United States in 1998 (Peyton et a1. 1999).

Systematics

The subgenus Ochlerotatus was elevated to genus level by Reinert in 2000 based on
morphological evidence. Prior to this re-classification, the genus Aedes was divided into
two subgenera, Aedes and Ochlerotatus. It has subsequently been suggested that
additional Aedini subgenera be elevated, including the subgenus Stegomyia (Reinert
2004). The most visible ramification arising from this change is the subgera elevation of
two medically-important species, the Yellow Fever mosquito, Aedes aegypti, and the

Asian Tiger mosquito, Aedes albopictus (Reinert 2004). As a result, elevation of these

genera has been a controversial subject and the complete adoption of Reinert’s
classifications by entomologists remains undetermined.

Medical Significance.

In addition being a nuisance to humans and animals, female mosquitoes ofien
transmit diseases. The range of maladies vectored by mosquitoes is extensive and the
importance of mosquitoes as disease vectors cannot be over-stated. For the purpose of
providing a background for this project, however, the discussion of mosquito-borne
diseases will be limited to those associated with A. albopictus, A. aegypti, and 0c.
triseriatus in their North American range, specifically arthropod-vectored viruses, or
arboviruses.

Of the arboviruses, yellow fever and dengue are the most serious and have had the
greatest historical impact on the field of medical entomology in the United States. Both
are associated with the Yellow Fever mosquito, A. aegypti, but A. albopictus is also a
competent vector of these arboviruses (Gubler and Rosen 1976). Outbreaks of dengue
and yellow fever were frequent in the United States from approximately the mid-
seventeenth century until the mid-twentieth century. (Bryan et al. 2004). Eradication of
larval habitats (i.e. standing water-filled containers), window-screening, and increased
pesticide use contributed to the near-eradication of US. dengue and Yellow Fever
outbreaks.

West Nile Virus (WNV) is the most visible mosquito-borne arbovirus,
particularly in recent years. Initially occurring during late summer 1999 in New York
City, WNV has subsequently spread across the continental US. to California (CDC 1999

a,b; Lanciotti 1999). Although mosquitoes in the genus Culex are the primary vector of

WNV, many Aedes/Ocherotatus species also transmit the disease, including 0c.
triseriatus and 0c. japonicus (Sardelis et al. 2001,2002; Turrell et al. 2005).

Mosquitoes that act as bridge vectors between birds and humans, such as C.
pipiens and C. restuans, are most likely to transmit WNV as birds serve as the reservoir.
Mammals, including humans, are generally unable to infect mosquitoes, due to low
viremia level in these hosts. Hence, Aedes/Ochlerotatus mosquitoes feeding on mammals
are less likely to account for a large percentage of positive mosquito pools. Culex
mosquitoes more strongly prefer bird hosts than do Aedes mosquitoes; therefore, Culex
are more likely to serve as a bridge vector of WNV between vertebrate and avian disease
hosts (Turrell et al. 2005).

Although not a major vector of WNV, 0c. triseriatus is the primary vector of La
Crosse Virus (LaCV), a California serogroup bunyavirus (Centers for Disease Control
and Prevention 2007, Watts et al. 1972). Although less common, A. albopictus has also
vectors LaCV in North Carolina and Tennessee (Gerhardt et al. 2001). LaCV is
distributed throughout the Eastern United States; it cycles between the mosquito vector
and vertebrate host (primarily chipmunks). Symptoms include encephalitis and, rarely
seizures and coma (Calisher 1983). Fatalities resulting from this disease occur in less than
1% of all clinical cases. Children are the primary risk group for LaCV.
Distribution and Life History

0c. triseriatus are opportunistic; they lay eggs in both natural and artificial
containers. In North America, Ochlerotatus triseriatus (Say) is the predominant mosquito
species in basal pan-type tree holes, but may also occur in man-made containers such as

tires, flower pots, and cemetery vases. The range of 0c. triseriatus encompasses much of

the Eastern United States and the southemmost sections of Eastern Canada, extending
from New Brunswick, Canada in the north to Florida in the south.

In northern latitudes such as Michigan, 0c. triseriatus overwinter as eggs within
their container habitat and emerge as larvae the following May, or earlier under warm
temperatures. Adult females lay individual eggs along the water line of natural and
artificial containers, where they may later be exposed to flooding during rain events.
Eggs are responsive to decreases in C02 and elevated moisture, utilizing these stimuli as
hatching cues (Clements 1992). The species is multivoltine contingent upon how early
adults emerge and whether late-season weather conditions remain favorable. Eggs
generally enter diapause by late September or early October.

Larval Habitat and Nutrition

In heterotrophic environments such as tree holes, nutrient inputs occur in three
main forms: as stemflow (water run-off from trees associated with rain events), animal
detritus, and allochthonous plant detritus. Leaves appear to be the most abundant source
of dissolved organic matter (DOM) and fine particulate organic matter (FPOM) to the
tree hole system. Animal detritus provides a richer source of nutrients to tree holes (Yee
and J uliano 2006), although it represents a comparatively smaller portion of the total and
the mechanism for its effect of increasing mosquito productivity has not been determined.
Leaf quality and quantity are important determinants of adult mosquito production (Fish
and Carpenter 1982, Leonard and Juliano 1995, Walker 1997). Leaf quality varies with
leaf type. Specifically, leaves with high tannin content, such as oak leaves, contain a
larger proportion of refractile (insoluable) to labile (soluble) material compared with

those lower in tannins, such as beech or maple (Walker 1997). For all leaf types,

senescent leaves are inferior to fresh leaves, possessing comparatively reduced
concentrations of soluble protein, nitrogen, and soluble carbohydrate. The quantity of leaf
detritus also conditionally influences the mosquito productivity (Leonard and Juliano
1995). In addition to determining adult size, the ration of leaf litter available per larva is
critical for larvae to enter the pupal stage. Available leaf rations interacted with several
variables in the tree hole, including larval density (interspecific and intraspecific),
presence of other macroinvertebrates, and stemflow.

Adult mosquito production in tree hole habitats is dependent on larval nutrition
derived from allochthonous nutrient inputs including leaf and animal detritus and
stemflow runoff. Microbial processing of these inputs increases the availability of detrital
nutrients to higher trophic levels in tree hole communities, which may also include a
diversity of macroinvertebrates in addition to mosquitoes (Carpenter 1983, Kitching
2001). Although mosquito larvae may ingest these detrital inputs in the form of fine
particulate organic matter (FPOM), it is more evident that the nutrients and physical
substrates provided by these inputs promote microbial colonization. Larval mosquitoes
obtain nutrition by browsing the microbial biofihn associated with leaf and container wall
surfaces or filter feed on small particles and planktonic bacteria in the water column
(Menitt et al. 1992). While bacteria are a critical resource for larval maintenance, it is
apparent that fungi supply larvae with additional nutrients required for growth (Kaufinan
et al. 2001 , 2002, Kaufman and Walker 2006).

Stemflow and detritus quality drive the production of adult mosquitoes, as these
factors directly impact microbial growth. Water running down tree trunks collects in tree

holes during rain events thus introducing critical soluble nutrients (e. g. nitrogen and

phosphorus and other inorganic nutrients) and carbon to the nutrient-poor ecosystem
(Carpenter 1982b). Stemflow provides critical nutrients need to promote decomposition
of detrital inputs. Microbial processing releases nutrients contained in leaf material to the
environment. The labile portioniof leaf material (= leachate), containing carbon and
nutrients (e. g. N and P) stored in leaf material, is released early in the decay process
approximately within three days of entering the system (Carpenter 1982a). The refractile
material, high in carbon but low in N and P, is largely unavailable to mosquito larvae. In
addition to solublizing the labile component of leaf detritus, stemflow promotes microbial
growth through introduction of limited nutrients. Microbes facilitate leaf decay by
sofiening tissues so they become available for direct ingestion by larvae. More
importantly, the microbial milieu, consisting of a diverse community of bacteria, fungi,
and protozoans, incorporate the nutrients from the organic inputs into their biomass.
Water column-associated materials as stated above are limited, but contain a higher
density of protozoans, if present, than leaf and container wall surfaces. While a superior
resource in terms of overall biomass for larvae compared with bacteria and fungi, these
organisms are rapidly depleted under larval feeding pressure (Kaufman 1999, 2001).
Ecological Factors Contributing to Adult Mosquito Production

In addition to the direct effect of stemflow, detrital inputs and microbial
processing, larval productivity is also affected by at least two other important indirect
factors: competition and facilitation. Although effects of these interactions on the
dynamics of tree hole macroinvertabrates has been the subject of several studies, the
effect of such interactions on the microbial community has remained relatively

unexplored. Given that the microbial community is the food resource driving competitive

interactions, one can predict that changes in the microbial community are likely to result
as changes occur in the structure of the community exerting feeding pressure on it. Such
interactions are likely to alter the microbial community, subsequently affecting adult
mosquito production. This “bottom-up” effect would result if changes in the microbial
community wrought by one species could reduce the relevant portion of the microbial
community to the lowest level necessary for the own survival but below the equilibrium
resource abundance (R*) threshold required for the competing species to maintain the its
own population density, that is, where birth rate is equivalent to death rate (1980, 1990).
This theory of competition has been described as a mechanism for organisms under
competition, but has remained unexplored as a mechanism directing the outcome of
competition among mosquito larvae. Rather, previous studies have focused on ration and
quality of detrital inputs experienced by competing species while omitting a description
of any explicit mechanism that would account for superior resource utilization.
Alterations in the structure or composition of tree hole associated microbiota are
important to our understanding of tree hole ecology, as they may translate into a positive
or negative effect on mosquito production.
Research Objectives and Rationale

The goal of this dissertation was to increase our understanding of the microbial
community resource base present in tree hole containers utilizing molecular genetic
techniques that promote analysis of community structure. Although the importance of
resource quality and quantity to mosquito development has been demonstrated, our
understanding how detrital resource inputs interact with microorganisms is limited;

therefore, in addition to understanding how resources contribute to mosquito success,

these studies seek to integrate microbial community level dynamics with broader
ecological processes associated with tree hole communities, including decomposition,
macroinvertebrate competition, and facilitation. Specifically, my dissertation objectives
were to:

1. Determine the effects of detrital leaching on mosquito productivity, including a)
whether the nutrients in leached senescent leaves are sufficient for growth; b) if
the positive effect of unleached leaves on mosquito development can be restored
by returning labile leaf components to microcosms; 0) whether fresh leachate
stimulates bacterial abundance and/or productivity; and (1) whether leachate alone
can support larval development.

2. Analyze the structure and diversity of microbial communities associated with
container habitats via terminal restriction fragment length polymorphism (T-
RFLP) analysis and sequencing of 16S and 18S rRNA genes. Included in this
objective was to: a) assess differences in the microbial community structure of
natural (tree hole) and artificial (tire) container habitats; and b) identify the
contribution of container substrates to mosquito production.

3. Investigate the microbial resource relationship with facilitation of 0c. triseriatus
populations by scirtid beetles using T-RF LP analysis of bacterial and fimgal
communities.

4. Investigate the microbial dynamics underlying competition between 0c.
triseriatus and A. albopictus and A. aegypti and A. albopictus using T-RFLP

analysis of bacterial and firngal communities.

CHAPTER 2
SENESCENT LEAF EXUDATE DECREASES MOSQUITO SURVIVAL IN TREE

HOLE HABITATS

10

Introduction

Tree holes, a type of phytotehnata (plant-based water-filled container) (Frank and
Lounibos, 1983), are small heterotrophic habitats harboring a diverse community of
macroinvertebrates and microorganisms (Carpenter, 1983; Kitching, 2001). In Eastern
North America, Ochlerotatus triseriatus (Say) is the predominant mosquito species and
primary macroinvertebrate consumer in basal pan-type tree holes (Craig, 1983). Mosquito
production in tree hole habitats depends on larval nutrition derived from allochthonous
detrital inputs. These inputs consist primarily of plant material, with leaf litter as the
major source of coarse particulate organic matter (CPOM), fine particulate organic matter
(F POM), and dissolved organic matter (DOM) in tree hole systems. In addition to leaf
detritus, stemflow and animal detritus also contribute to allochthonous nutrient pools.
Water runoff during rain events collects in tree holes, introducing critical soluble
nutrients (e. g. nitrogen, phosphorus and other inorganic nutrients) and carbon to the
nutrient-poor ecosystem (Kitching, 1971; Fish 1983; Carpenter, 1982b; Walker and
Merritt, 1991). Macroinvertebrate carcasses and fecal material also contribute to the
detrital pools (Daugherty et al., 2000; Yee and Juliano, 2006), although the presence of
these materials is comparatively ephemeral.

Although mosquito larvae may ingest detrital inputs in the form of F POM, it is
more evident that the nutrients and physical substrates provided by these inputs promote
microbial colonization; thus, adult production is indirectly linked to detrital inputs
(Kaufinan and Walker, 2007). Analyses of larval gut content and feeding behavior
corroboratively indicate mosquitoes obtain food by browsing the microbial biofilrn

associated with leaf and container wall surfaces or by filtering small particles such as

11

planktonic bacteria from the water column (Cummins and Klug, 1979; Fish and
Carpenter, 1982; Merritt et al., 1992; Walker and Merritt, 1991). The tree hole-associated
microbial milieu consists of a diverse community of heterotrophic bacteria, fungi, and
protozoans (Kaufinan et al., 2001, 2002; Kaufman and Walker 2007). These
microorganisms play an important role in tree hole ecosystems, contributing to carbon
and nutrient cycles through the secondary production of microbial biomass and the
recycling of organic carbon and nutrients thus increasing the availability of detritus-
derived nutrients to higher trophic levels in tree hole communities (Kaufman and Walker
2006)

Adult mosquito production is conditionally dependent on the available per larva
leaf ration (Fish and Carpenter, 1982; Hard et al., 1989 Leonard and Juliano, 1995;
Walker et al., 1997) and the interaction thereof with habitat variables, including larval
density, presence of other macroinvertebrates, and stemflow. The quality of leave
material is also of critical importance, indicated in part by observations of greater
mosquito production fi'om microcosms stocked with fresh leaves compared with
senescent leaves (Walker et al., 1997). Additionally, qualitative differences occur among
leaf litter types such that species with faster decomposition rates are generally superior,
supporting greater mosquito growth and survival than those with slow decomposition
rates (Fish and Carpenter 1982). Decomposition of leaf material is associated with its
availability for microbial degradation and its palatability for macroinvertebrates, as both
processes are governed by the presence of lignin and nitrogen content (Barlacher 1985,
Moorehead and Sinsabaugh, 2006). Indeed, the presence of carbohydrates and other

nutrients, particularly N, has a positive effect on the fungal productivity and mosquito

12

growth parameters. (Kaufman and Walker 2006) For all leaf types, senescent leaves are
inferior to fresh leaves, possessing comparatively reduced concentrations of soluble
protein, nitrogen, and soluble carbohydrate.

Leaf decomposition occurs in two distinct stages: initial leaching of labile
components over a short time period and a long-term breakdown of refractory
components (Carpenter, 1982a), Release of soluble leaf components occurs early in the
decay process, approximately within 24 hours of entering the system (Gessner and
Schwoerbel 1989, Webster and Benfield 1986). C:N ratios vary among leaf species,
resulting in relative differences in the portions of labile and refractile leaf components
and subsequent, differential breakdown among leaf types. The labile material, or
leachate, is rich in nutrients that are critical for microbial and mosquito productivity.
Refractile material, on the other hand, has a high C:N ratio, low phosphorus content and
is likely unavailable directly to developing larvae (Carpenter, 1982a; Webster and
Benfield, 1986). Microorganisms afford a bridge between the nutrients trapped in the leaf
matrix and mosquito larvae by utilizing leaf material as substrates for growth while
contributing to leaf breakdown.

The soluble leaf fraction represents the most important contribution of nutrients
to larval productivity, yet this material disappears rapidly from the leaf matrix. Thus,
leaves that are subjected to leaching prior to entering the tree hole may be of lower
quality than unleached leaves and may physically impede inputs of higher quality leaves.
Leaching has dampening effects on the growth responses of mosquitoes due to the
consequent reduction in nutrient content (Walker et al., 1997). It follows that re-addition

of leached contents should restore the required nutrients to systems stocked. with leached

l3

leaves. Furthermore, the effect of high nutrient concentration should manifest itself via
the increased production of microorganisms. This prediction follows from the results of
previous studies showing enhanced mosquito growth in response to nutrient
supplementation via the microbial loop (Kaufman et al., 2002; Kaufman and Walker,
2006).; therefore, additions of leachate material obtained from an equivalent leaf pack
mass should mitigate the effects of poor leaf quality on the development and productivity
of Oc. triseriatus.

We describe here three experiments designed to elucidate the individual effects of
labile and refractile leaf components on tree hole community dynamics. The purpose of
this study was to determine: 1) whether the nutrients in leached senescent leaves are
sufficient for growth; 2) if the positive effect of unleached leaves on mosquito
development can be restored by returning labile leaf components to microcosms; 3)
whether fresh leachate stimulates bacterial abundance and/or productivity; and 4) whether
leachate alone can support larval development. We postulate that leaf litter present in tree
holes contributes relatively little to mosquito growth compared to fresh inputs of leaf
material; therefore, leached leaves alone should be insufficient for larval growth. A
corollary of this hypothesis is that the leached fiaction, containing labile substrates and
nutrients to initiate high microbial activity, will support larval mosquito grth and
development comparable to that observed in response to fresh leaf packs. Additions of
leachate material obtained from an equivalent leaf pack mass should therefore “rescue”
larvae from the effects of poor leaf quality. Further, we predict that bacterial populations

will be enhanced in response to leachate due to the high nutrient content (Kaufman et al.,

14

2002; Kaufman and Walker, 2006) compared with populations that experience reduced
leachate.
Materials and Methods

Microcosm construction

For each of the following experiments, tree hole-based microcosms were stocked
with senescent red oak leaves (Quercus rubra L.) collected at Michigan State
University’s Kellogg Forest (Augusta, MI). Leaves were dried at 45°C for 48 h and added
as l-g leaf packs to microcosms constructed as in (Kaufman and Walker, 2006; Walker et
al., 1991). Additionally, microcosms received a microbial inoculum, consisting of 3 m1
homogenized natural tree hole water and particulates. Each contained a final volume of
500 ml, composed of deionized, distilled water and, if applicable, leachate in the amounts
described below. Water levels were maintained throughout the experiment to account for
evaporative losses. Microcosms were loosely covered with window screen and incubated
under indirect lighting at 21 °C and 16:8 (LzD) photocycle (Percival Scientific, Inc.,
Perry, IA). Prior to the addition of 20 or 40 newly-hatched first instar 0c. triseriatus
larvae (Day 0 for all experiments), microcosms were incubated for 3 days to allow time
for microbial colonization of leaf surfaces and water column. The larvae used in the
following experiments were hatched from eggs collected from our colonies at the Insect
Microbiology Laboratory at Michigan State University.
Experiment 1.

The purpose of this study was to elucidate the contribution of the labile fraction of
senescent leaves as detrital inputs into microcosms that model tree hole habitats. Leaf

quality and leachate effects were assessed in a 2 x 4 multifactorial design with six

15

replicates per treatment. Unleached senescent red oak leaf (Quercus rubra) packs (1 g)
were compared to similar leaves subjected to leaching for three days. This period is
sufficient to account for the leaching of the labile components into the water column as
the majority of this fraction is lost fiorn the leaf matrix within three days of introduction
to an aquatic environment (Carpenter, 1982a). Microcosms containing newly hatched
first instar 0c. triserz'atus larvae, described above, received the resulting leachate in
amounts equivalent to 0, 25, 50, or 100% of that obtained from a 1 g leaf pack. To
account for the effect of labile nutrients alone on mosquito performance, all leachate was
filter-sterilized using a 0.2 pm vacuum filter before addition to microcosms. Microcosms
were checked daily for adult mosquitoes, which were collected and stored at -80°C. At
the end of the experiment (day 70), adults and any remaining larvae and pupae were
lyophilized and massed. The remaining leaf mass was also determined after drying leaves
for three days at 50°C.

Experiment 2.

We tested the hypothesis that soluble nutrients and microbial biota present in the
leachate fi'action of leaf material increase mosquito performance in microcosms similar to
those described above. Two levels of leaf quality, unleached or leached for three days,
were applied to replicate microcosms. In contrast to the previous experiment, leachate
was added to microcosms equivalent to 100% of the amount obtained from a 1 g leaf
pack in two forms: unfiltered or pro-filtered through a 0.2 pm vacuum filter. Additionally,
a control treatment of deionized, distilled water was applied to replicate microcosms for
each leaf quality treatment. The resulting 2x2x2 factorial design was replicated seven

times to permit the replacement of leaf material sampled from six replicate microcosms

16

on day 0 (prior to the addition of larvae). At day -3, filtered and unfiltered bulk leachate
were sampled for bacterial productivity, bacterial abundance, and nutrient analysis.
Microcosm sampling was done at the onset and at the termination of the experiment (days
0 and 70, respectively). Additional nutrient samples were taken from microcosms several
weeks into the experiment (day 12). Water samples were collected for bacterial
productivity, bacterial abundance, and nutrient analysis (1-10 ml). Productivity
subsamples (1 ml) were maintained at 20°C and abundance subsamples (5 ml) were
preserved with formaldehyde at a final concentration of 3.7% until measurements could
be taken. Leaf material was subsampled using a cork borer (10 mm diam) for estimates of
bacterial abundance. Two discs were aseptically removed from leaf packs into 5 ml filter-
sterilized phosphate buffer and preserved with formaldehyde (3.7% final concentration).
Larvae, pupae, adults, and leaf packs were treated as described above to obtain dry
weights.

Experiment 3.

In this experiment, we compared the relative contributions of labile and refractile
leaf components to mosquito productivity. Senescent red oak leaf packs (1.0g) were
leached for three days in 12 replicate microcosms containing 500 ml deionized, distilled
water. The water from six microcosms, now containing labile leaf components, was
poured into new microcosms and replaced with fresh deionized, distilled water. The
resulting treatment design was thus: leached leaf + water, leached leaf + leachate, no leaf
+ leachate. As in the previous experiments, mosquitoes at all stages of development were
collected and processed to obtain dry mass measurements.

Chemical analyses

17

Total nitrogen (N) and phOsphorus (P) present in the samples were quantified via
spectroscopy of unfiltered water samples (Kaufman and Walker 2006). For each analysis,
persulfate oxidation techniques were used to convert all forms of phosphorus and
nitrogen to phosphate and nitrate, respectively (Menzel and Corwin 1965, Crumpton et
al. 1992, Bachmann and Canfield 1996). A colorimetric assay was used to enumerate
total P (Murphy and Riley 1962), while second derivative spectroscopy was employed for
enumeration of total N (Crumpton et al. 1992, Bachmann and Canfield 1996).

Bacterial abundance.

Bacterial abundance on the leaf surface and in the water column sub-samples was
quantified via direct microscopic counts using the DAPI (49,6-diamidino-2-phenylindole)
fluorescent staining procedure (Porter and Feig 1980, Walker et al. 1988, Kaufman et al.
2001). Water column and leaf disc samples were sonicated (Aquasonic model SOT,
Westchester, PA) for 12 min to reduced cell clumping and/or dislodge cells (Velgi and
Albright, 1993). Samples were vortexed and diluted as necessary with filtered-sterilized
Milli-Q water (Millipore, Bedford, MA). After the staining material at a final
concentration of 20 ug/ml for 15 min, samples were filtered onto black filters (0.2-mm
pore size; Nucleopore, Costar, Cambridge, Mass). For each subsample, filters were
counted (600 cells per filter minimum) at 1000X.

Bacterial productivity

Direct measurements of microbial biomass accumulation were conducted using a
3H-leucine incorporation assay (Kirchman 2001). This technique measures of the
incorporation of amino acids into protein in a bacteria-specific manner, through the use of

short incubation periods and nanomolar leucine concentrations (Riemann and Azam

18

1992). A 5.85 ratio of labeledzunlabeled leucine was added to water subsamples at a
concentration of 25 nM to achieve saturation of uptake kinetics (Kirchman 2001,
Kaufman et al. 2001). Water samples were incubated with labeled leucine (L-leucine
(4,5-3H ), 50 Ci/mmol- NEN, Life Science, Boston, MA) in the dark at room temperature
for 30 min in 2 ml microcentrifuge tubes (Smith and Azam 1992, Kirchman 2001).
Trichloroacetate [TCA, final concentration 10% (vol:vol)] was added to terminate
reactions and precipitate protein. Two rinses of the TCA-protein precipitates were
conducted with 10% TCA, followed by a single rinse with 5°C, 80% (vol:vol) ethanol.
Standard liquid scintillation counting techniques were used to quantify the amount of
radioactivity present in the samples.
Statistical Analysis

Within each experiment, multivariate analysis of variance (MANOVA)
techniques were used to analyze groups of related variables (Proc GLM, SAS Institute).
Specifically, all mosquito parameters measured, (total survival, male and female mass,
and male and female development time) were grouped within a single MANOVA.
Dependent variables with significant MANOVA results were subjected to individual
univariate analysis of variance (AN OVA) followed by a Bonferroni correction to reduce
the chance of Type 1 error (Rice, 1989). Bacterial parameters and nutrient concentrations
measured in experiment 2 were analyzed using separate repeated measures MAN OVAs
(also called doubly-multivariate repeated measures MANOVA) to account for the effect
of time on measurements. These variables were grouped separately from mosquito
variables because they respond differently to treatment combinations (Kaufman et al.,

2002). For the same reason, remaining leaf mass was analyzed independently of other

19

dependent variables in a separate AN OVA for each experiment. When necessary, data

were square-root [(x + 0.5)”2

] or arcsine transformed prior to analysis to meet normality
criteria. All values reported are non-transformed. Following univariate analysis, means
separation was performed for significant independent variables in experiment 3 using
Fisher’s protected least significant difference (LSD) test.

Results
Experiment I .

The main effects of leaf type and leachate had significant effects on mosquito
production parameters. A significant interaction effect between leaf and leachate in this
multivariate analysis of variance suggests that leaf effects changed with increasing
amounts of leachate added to microcosms (Table 2.1). Mosquitoes in microcosms
containing unleached leaves were characterized by significantly increased survival,
reduced development time, and increased mass of male adult 0c. triseriatus (univariate
analyses, Figure 2.1, Table 2.1), indicating that unleached leaves were a better resource
for mosquitoes. In contrast, univariate analyses showed leachate additions were
associated with significant changes in development time and body mass parameters for
male mosquitoes only, with leachate additions of 100% producing increases in the former
condition and reductions in the latter compared with the other treatments (Figure 2.1,
Table 2.1). In all cases, mosquito parameters associated with additions of 100% leachate
to leached leaves did not recover the level of those parameters associated with additions
of 0% leachate to unleached leaves such that survival and adult body mass obtained were

comparatively lower and development time was slower. Furthermore, additions of

leachate to unleached leaves generally did not affect mosquito growth. Female mass was

20

the single exception to this observation; greater mass accrued in females provided with
unleached leaves and 100% leachate compared with all other treatments. Finally, a
significantly greater amount of leaf mass was lost over the course of the experiment from
the leaves that were not subjected to leaching prior to their addition in microcosms

(F=5.l3; df=7,39; p=0.0003; Figure 2).

21

Table 2.1. Multivariate (MANOVA) and univariate (AN OVA) hypothesis test results for
all factors and their interactions for mosquito production variables in experiment 1.
ANOVA results for each variable are shown for factors with significant MANOVA

(P<0.05).

 

Wilks’ Response variable
Source Lamda tested with AN OVA F (If P
Leaf 0.117 27.17 5,18 <0.0001*
Female mass 3.00 1,22 0.097
Male mass 4.41 1,22 0.047
Egrgale development 13.27 1,22 0.001 *
Male development time 58.89 1,22 <0.0001*
Survival 1 1.2 1,22 0003*
Leachate 0.108 4.15 15,50 <0.0001"‘
Female mass 1.99 3,22 0.145
Male mass 3.67 3,22 0.028
Efnrréale development 129 3,22 0301
Male development time 6.92 3,22 <0.0001*
Survival 0.47 3,22 0.706
Leaf‘Leachate 0.179 4.91 10,36 0002*
Female mass 2.62 2,22 0.095
Male mass 4.09 2,22 0.031
rhinerale development 0.26 222 0772
Male development time 15.58 2,22 <0.0001*
Survival 5.57 2,22 0.011

 

* Indicates significance at a-value < 0.05 after sequential Bonferroni correction.

22

 

 

 

 

 

 

 

 

 

 

 

 

 

    

 

 

 

 

 

   

$3 100 A
'5
E El Leached leaves
3
0:3 I Unleached leaves
(0
(D
E
d)
.e 601 C
E 50‘
‘é’ 4o .
8'”; l 40
7;; .8 30 30‘
8 2o 20*
I:
3; 10 ml
2
0 l..." ,,_ O W raj
3, 0-5lD 05.
E l l
a, 0.4] 0.4!
g .
E 0.3 0.31
% 0.2: 0.2:
(U
E 0.1‘ 0.1 I I
(D
2 O -_ O 4
0 1o 50 100 0 1o 50 100
Leachate Added (%)

Figure 2.1. Mosquito production variables from experiment 1. (A) Survival (20 initial
larvae). (B) Average female development time. (C) Average male development time. (D)
Average female weight. (E) Average male weight. Values are means d: SE (n = 6 for all

variables in A, C, and E. n = 2-6 for variables in B and D).

23

 

900

3
E
or 800 ~
.E
.E
g 700 a
9
3
m 600 w
E
“(B
a) 500 -
_.|

400 ~

 

 

 

 

 

 

 

 

 

 

 

o 10 50 100
' Leachate Added (%)

Figure 2.2. Leaf mass lost in experiment 1. Values are means :b SE (n = 6).

24

 

Experiment 2.

As in the first experiment, unleached leaves were a better resource for
mosquitoes, supporting significantly greater adult production in these microcosms
compared with microcosms containing unleached leaves (Table 2.2, Figure 2.3). Adult
mass was greater and development time reduced by the presence of unleached leaves.
Similarly, additions of leachate to microcosms had a significantly positive affect on
mosquito growth. Unlike the first experiment, the effect of leached leaves on mosquito
production was mitigated by the addition of leachate such that adult emergence in
microcosms with this treatment combination was equal to or greater than emergence in
microcosms with unleached leaves and 0% leachate. The production of adult mosquitoes
was not effected by the condition of the leachate added (e. g. filter-sterilized or non-
filtered) (Figure 2.3). Furthermore, filtration did not significantly interact with other

treatment combinations to affect mosquito growth parameters under any condition.

25

Table 2.2. Multivariate (MAN OVA) and univariate (ANOVA) hypothesis test results for
all factors and their interactions for mosquito production variables in experiment 2.

AN OVA results for each variable are shown for factors with significant MANOVA
(P<0.05).

 

Wilks’ Response variable
Sources Lamda tested with AN OVA F (If P
Leaf 0.326 7.85 5,19 0.0004*
Female mass 0.15 1,23 0.704
Male mass 0.66 1,23 0.425
firenrréale development 005 1’23 0.81 9
Male development time 0.76 1,23 0.392
Survival 6.76 1,23 0016*
Leachate 0.16 19.93 5,19 <0.0001*
Female mass 4.37 1,23 0048*
Male mass 11.97 1,23 0002*
Elenrrcrale development 3.13 1,23 0.09
Male development time 15.33 1,23 0001*
Survival 18.52 1,23 0.0003*
Filter 0.66 1.95 5,19 0.132
Leaf x Leachate 0.588 2.66 5,19 0.055
Leafx Filter 0.805 0.92 5,19 0.488
Leachate x Filter 0.853 2.54 5,19 0.063
Leaf x Leachate 0.853 0.66 5,19 0.66

x Filter
* Indicates significance at a-value < 0.05 after sequential Bonferroni correction.

26

Table 2.3. Experiment 2 multivariate (MANOVA) and univariate (AN OVA) hypothesis
test results for all factors and their interactions for productivity and abundance of bacteria

on leaf surfaces and in the water column. ANOVAs were performed for significant
ANOVA factors (P<0.05).

A) Between subjects

Sources MS F df P
Leaf 18.4 68.3 1,23 <0.001*
Leachate 2.1 7.6 1,23 001*
Leaf x Leachate 0.6 2.4 1,23 0.14
Filter 2.0 7.3 1,23 0.01*
Leafx Filter 5.2 19.3 1,23 <0.001*
Leachate x Filter 2-7 9-9 1,23 001*
Leaf x Leachate x Filter 2-3 8'5 1,23 001*
B) Within subjects Wilks’

Sources Lamda F df P
Time 0.1 108.3 3,21 <0.001*
Leaf x Time 0.1 43.4 3,21 <0.001*
Leachate x Time 0.4 11.8 3,21 <0.001*
Leaf x Leachate x Time 0.5 8.4 3,21 <0.001*
Filter x Time 0.5 8.4 3,21 <0.001*
Leaf x Filter x Time 0.4 8.6 3,21 <0.001*
Leachate x Filter x Time 0.4 9.6 3,21 <0.001*
Leaf x Leachate x Filter x Time 0.5 8.0 3,21 0001*

 

* Indicates significance at a—value < 0.05 after sequential Bonferroni correction.

27

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

   

 

 

 

 

80
§ 70 A
{—D 60
> .
'E 50 El Filtered leachate
(1:) 40
c 30 I Non-filtered leachate
m
g 20
1O
0 .
60 60
E B c
(D 50‘ 50.
E or
.2 5 4° ..
0 'c
a 7.; .. ..
1’ E
g z; 20 20
m
2 10 10
O
A 04 0.4
E, D E
a 0.3 03
(B
E
+- 0.2 0.2
3
1:
g 0.1 0.1l
cu
m
2 o . o r
Not Not Not Not
Leached leached Leached leached Leached leached Leached leached
No Leachate Leachate No Leachate Leachate

 

Figure 2.3. Mosquito production variables from experiment 2. (A) Survival (40 initial
larvae). (B) Average female development time. (C) Average male development time. (D)
Average female weight. (E) Average male weight. Values are means :1: SE (n = 6 for all
variables in A, C, and E. n = 1-6 for variables in B and D).

Microbial parameters. Bacterial abundance and productivity were significantly

affected by leaf condition, leachate type, and filtration of the leachate, as indicated by

28

MANOVA with repeated measures on these parameters (Figures. 2.4 and 2.5; Table 2.3).
Significant interactions were detected among all main effects, with the exception of leaf
and leachate. Productivity of bacteria in the water column was greatest in response to
unleached leaves at day 0 (prior to larval addition) when unleached leaves were present.
Moreover, productivity was higher for this treatment combination in the presence of
filtered leachate.

In addition, time had a significant affect on bacterial productivity and abundance.
This trend disappeared by day 70 of the experiment, however. In general, bacterial
productivity dropped below 1 x 10-7 pmol/ml for all treatment combinations at this time
(Figure 2.4), although direct microscopic counts indicate that bacteria remained present at
a similar abundance over the course of the experiment (Figure 2.5). Bacterial abundance
in the water column responded in a similar manner to treatment combinations, with
significantly higher concentrations of cells/ml at day 0 evident in microcosms receiving
unleached leaves and filtered leachate (Figure 2.5; Table 2.3). By day 70, water column

productivity had dropped to statistically similar lows for all treatments.

29

El Filtered I Non-filtered

 

 

4 DAY 0

 

 

 

 

 

 

 

 

 

 

 

 

pmol/ml/h

4 DAY 70

 

 

 

0 Mai

Not Not
Leached leached Leached leached

No
Leachate Leachate

Figure 2.4. Water column bacterial productivity (leucine incorporation rate) response to
leachate, filtration, and leaf condition in experiment 2. (A) day 0 and (B) day 70. Values

are means :1: SE (n = 6). Values are means :1: SE (n = 6).

30

DAY 0 DAY 70

 

 

7.6 A 3
7.2 7
6.8
6.4

6.0
5.6

WATER
COLUMN

Log cells/ml

   

5.2
4.8 -
7.6
7.2

 

 

 

 

 

 

 

 

 

     

 

 

 

 

 

6.8

LEAF
SURFACE

6.4 l

6.0
5.6

Log cells/disc

       

5.2

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

4.8

Not Not Not Not
Leached Inch“ Leached med Leadred leached Leached law
No Leachate Leachate No Leachate Leachate

D Filtered leachate I Non-filtered leachate

Figure 2.5. Bacterial abundance (direct microscopic counts) response to leachate,
filtration, and leaf condition in experiment 2. Water column: (A) day 0 and (B) day 70.
Leaf surface: (C) day 0 and (D) day 70. Values are means :t SE (n = 6).

31

Nutrient Analysis. Total N and P concentrations (Figure 2.6) were analyzed using
a 2 (leaf) x 2 (leachate) x 2 (filtration) doubly- multivariate repeated measures
MANOVA. This showed a significant main effect for leaf type and leachate, but no
significant main effect for filtration (Table 2.4). Microcosms containing leached leaves
contained greater amounts of N and P relative to those with unleached leaves. Similarly,
microcosms with leachate contained higher N and P levels compared with microcosms
lacking leachate (Figure 2.6).

Both leaf and leachate significantly interacted with filter, and interactions were
evident among leaf, leachate, and leaf x leachate with time (Table 2.4). Specifically, total
N and P concentrations decreased over time; at days 0 and 12, the content of N and P in
microcosms containing leached leaves and no additional leachate was lower than in
microcosms stocked with leached leaves and leachate. Follow-up univariate analyses

indicate both N and P contributed to these results.

32

Table 2.4. Multivariate (MANOVA) and univariate (AN OVA) hypothesis test results for

all factors and their interactions for nitrogen and phosphorus concentration in experiment

2. AN OVA’s were performed for significant AN OVA factors (P<0.05).

 

A) Between subjects

Sources MS F df P
Leaf 13.6 1,40 <0.001*
Leachate 8.3 l ,40 <0.001 *
Leaf x Leachate 0.03 1,40 0.22
Filter 0.0002 1 ,40 0.92
Leafx Filter 0.3 1,40 <0.001*
Leachate x Filter 0.1 1’40 0-02
Leaf x Leachate x Filter 0-01 1,40 0-47
B) Within subjects

Sources Wilks’ Lamda F (If P
Time 0.1 79.9 4,37 <0.001"‘
Leaf x Time 0.1 93.0 4,37 <0.001*
Leachate x Time 0.3 24.3 4,37 <0.001*
Leaf x Leachate x Time 0.5 7.7 4,37 <0.001*
Filter x Time 0.8 1.7 4,37 0.17
Leaf x Filter x Time 0.9 0.6 4,37 0.66
Leachate x Filter x Time 0.9 0.8 4,37 0.52
Leaf x Leachate x Filter x Time 0.9 0.6 4,37 0.65

 

* Indicates significance at a-value < 0.05 after sequential Bonferroni correction.

33

Figure 2.6. Water column chemistry for experiment 2 at days 0 (A,B), 12 (CD), and 70

(E,F). Shown are total nitrogen (A,C,E) and total phosphorous (B,D,F).

34

49m. 2 4.2m. n

8
>'
u:

w om.

 

oh 4.

a...

.. 0.».

.. o...

:Hll i .. -_ r . r
O.

A 9m.

Om<o

 

 

 

  

somamm

mglL

owaoroo
O
N

Om< AM

 

 

 

 

on.“
- oifll _ I I

d

O
m
m

 

Um< Mo

 

onamoo
O
N

l
l

IMHIIlH'l _

35

rmmosma

Zo.
.mmoaoa

romoyoa

20 5838

20»

_mmo:oa
romozmno

romosoa

zon
.mmozma

zO rmmosmnm

20»
_omo:oa

 

 

rmmozmno

3.62.5 .9336 l 203-3643 Emerge

Ems—d Pa.

Experiment 3.

Mosquito parameters were significantly affected by the type of input provided in

microcosms (Table 2.5). Compared with microcosms receiving leachate only,

microcosms containing leached and unleached leaf treatments produced significantly

more adult mosquitoes, and these developed faster and attained greater mass (Table 2.5,

Figure 2.7). Mosquito performance in leachate only microcosms was poor, with only two

males and zero females produced with this treatment.

Table 2.5. MANOVA results for all factors and their interactions for mosquito

production variables in experiment 3. AN OVA results for each variable are shown for

factors with significant MANOVA (P<0.05).

Response variable tested

 

Source Wilks’ Lamda with ANOVA F df P

Treatment 0.304 3.53 6,26 001*
Female mass 1.66 2,15 0.22
Male mass 5.03 2,15 0021*
Female development time 21.21 1,3 0019*
Male development time 5.29 2,7 004*
Survival 8.37 2,15 0004*

 

* Indicates significance at a—value < 0.05 after sequential Bonferroni correction.

36

 

 

900 .r 80

800 . a a

3 60
700» i
1 40
600

500 b ' 20

or
(%) lawns ueew

 

 

   

 

 

 

 

400

 

 

50

40

30

(days)

20‘

Mean development time Leaf mass remaining (mg)

10

 

 

 

 

 

 

 

 

0.8 .
0.6 .
0.4

0.2 .

Mean adult mass (mg)

 

 

 

 

   

 

Leachate Leached Unleached Leachate Leached Unleached

Treatment

Figure 2.7. Mosquito production variables from experiment 3. (A) Average leaf mass
remaining. (B) Survival (40 initial larvae). (C) Average female development time. (D)
Average male development time. (B) Average female weight. (F) Average male weight.
Values are means i SE (n = 6 for all variables in A, B, E, and F. n = 0-6 for variables in
C and D). Means within each column followed by the same letter are not significantly

different, (P > 0.05, Fisher’s Protected LSD Test).

37

Discussion

Senescent leaves are the most abundant allochthonous input in tree holes.
Previous work suggests that the labile fraction of leaf material is critical to mosquito
growth (Walker et al., 1997). In this experiment, mosquito production fell as much as
38% in microcosms treated with leached leaves. Thus, we expected that the positive
effect of labile components on population growth could be reclaimed by the introduction
of labile leaf components to microcosms containing leached leaves. In our first
experiment, mosquito survival (larvae and adults) in microcosms containing leached
leaves and 100% of the leachate produced by a an equivalent leaf pack was not
significantly different from survival in those containing unleached leaves and no
additional leachate, suggesting that for this parameter the effect of leachate could be
restored. In contrast, development time and adult mass for males and females did not
exhibit a positive response to the reintroduction of leachate. We postulate that this
divergent effect occurred because leached leaves were adequate for maintenance of
mosquitoes at the larval stage, but insufficient for driving adult production. Apparently,
the failure of leachate restoration to stimulate equivalent adult production was due to the
effect of filtering the leachate. A plausible explanation is that the available nutrients
present in the leached leaf material were assimilated by microorganisms prior to filter-
sterilization during the leaching period. Filtering the leachate before adding it to
microcosms would, therefore, remove two critical components from the microcosms:
incorporated nutrients and established populations of microorganisms.

Surprisingly, the results of Experiment 2 showed that filtration of leachate did not

influence mosquito production parameters despite an initial surge in bacterial abundance

38

and productivity in microcosms receiving filtered leachate and unleached leaves.
Evidently, microbial dynamics were affected by the treatment of leachate prior to its
addition in microcosms. Greater productivity was observed following filtration,
presumably due to the removal of established bacterial populations, allowing freshly
inoculated populations to enter into an exponential growth stage. That bacterial
populations were present prior to filtration is evidenced in experiment 1, wherein DMCs
of bacterial cells in leachate prior to filtration were 7.2 log cells/ml. In addition, removal
of protozoan predators may contribute to greater productivity following filtration.

The success of mosquitoes in the unleached leaf treatments compared with
leached leaf treatments (no leachate added) in experiments 1 and 2 and in Walker et al.
(1997) was presumably associated with corresponding high bacterial production and
abundance at day 0. These high levels were not maintained throughout experiment 2,
with differences among treatments becoming negligible by day 70. This result is not
surprising, given previous work which indicates that grazing pressure fi'om predators
reduces bacterial productivity (Kaufman, 1999, Findlay, 1986). It should, however, be
noted that the microbial analyses herein do not account for compositional changes in the
microbial community. Indeed, microbial communities exposed to invertebrate feeding
pressure are known to undergo a structural shift, potentially resulting in the dominance of

indigestible forms (J iirgens and Matz, 2002)). Hence, microbial abundance may remain

relatively unchanged throughout an experiment despite underlying changes in bacterial
susceptibility to foraging mosquito larvae. That declines in bacterial productivity do not
correspond with drops in abundance of the same magnitude also show that neither

community composition nor metabolic activity were altered. The absence of changes in

39

bacterial abundance also suggests sufficient nutritive material is available to maintain
microbial communities in the absence of stemflow events (which would bring in pulses
of limiting nutrients such as N and P).

The failure of leachate alone to support larval development (experiment 3)
suggests either the importance of surfaces, insufficient nutrient supply, or negative effects
associated with high tannin content in the leachate . Surfaces are important to
microorganisms, as they support the formation of biofilrns, promote production of fungi,
and have been shown to be grazed by larvae. Nutrients and DOC continue to leach from
the leaf matrix, although at a greatly reduced rate.

In nature, rapid degradation of leaves may occur upon entering tree holes,
particularly under high nutrient conditions (Macia and Bradshaw, 2000). Nutrients (N, P)
promote decomposition because, despite available carbon pools, the production of
microorganisms is limited by nutrients. Once the initial flush of nutrients from the leaf
matrix has been exhausted, heterotrophic microorganisms are nutrient limited until
replenishing stemflow or detritus enter the system. Although these initial fluxes of
nutrient-rich leachate are critical to developing larvae, they may not be the norm. More
likely the situation larvae experience is one supported by low quality leaf material. As
Oc. triseriatus larvae hatch from mid- to late spring, leaf inputs are primarily in the form
of senescent leaves that have been subjected to some degree of leaching throughout the

winter.

40

CHAPTER 3
HABITAT-SPECIFIC CHANGES IN MICROBIAL COMMUNITY DIVERSITY

ASSOCIATED WITH CONTAINER-BREEDING MOSQUITOES.

41

Introduction

Assessments of the microbial community composition and diversity are required
if we are to understand the firnction and relative importance of individual microbial
populations in tree holes and other container habitats, and to determine how these
populations may contribute to overall mosquito productivity. Biotic and abiotic factors
may impact the levels of richness and evenness in the microbial communities associated
with tree holes (Bell 2005). Conversely, the relative productivity of mosquitoes in tree
holes may vary directly in response to changes in the microbial community composition.
Biotic factors may include the presence/absence or abundance of macroinvertebrates
within tree holes, while abiotic factors include, but are not limited to detrital and nutrient
inputs, stemflow events, pH, temperature dissolved oxygen concentration, and habitat
size. The composition of leaf surface and water column-associated microbial
communities of tree holes has been previously assessed in our laboratory and others via
16S rDNA sequence analysis and other molecular genetic approaches, fatty acid methyl
ester (FAME) profile analysis (Kaufinan et al. 1999, 2008; Xu et al. 2008), and DGGE
(Bell 2005); however, container wall surfaces, an important component of the tree hole
habitat, have not been examined. Furthermore, the microbial community associated with
any component (water column, detrital surface, container surface) has remained
unexamined in artificial container habitats such as tires, in which many medically
important mosquito species commonly breed. Understanding the microbial dynamics of
these habitats is critical if we wish to compare tree hole and tire systems, an important
area of study given the preference of many medically important mosquito species (e. g. A.

albopictus and 0c. japonicus) for these habitats.

42

There are a number of molecular genetic approaches available to assess microbial
community composition that are culture independent (Dori go et al. 2005, Talbot et al.
2008). Among these, analysis of terminal restriction fragment length polymorphisms (T-
RF LPs) obtained from 16S and 188 rRNA genes is a relatively inexpensive method for
analyzing microbial communities (Liu et al. 1997, Marsh 1999). This method provides an
attractive alternative to traditional cloning and sequencing of microbial DNA due to the
reduced cost and high throughput associated with sample processing. The decreases in
time and expenditure result in microbial community analysis of replicated, manipulative
experiments being a feasible option. The method has been employed previously in studies
of bacterial and fungal communities in diverse habitats, but has not yet been used in
published studies of larval mosquito habitats (Maknojia 2006) Bacterial diversity and
nutritional significance of the surface microlayer in Anopheles gambiae (Diptera:
Culicidae) larval habitats.

Our objective in the current study was to assess differences in the microbial
community structure of tire and tree hole habitats. Specifically, we answered the
following questions: 1) How does microbial community diversity vary among associated
tree holes and tire habitats?; 2) does the microbial community of a container habitat
change throughout the season i.e. does community succession occur?; 3) do the microbial
diversity measurements obtained using t-RFLPs correspond to the measurements
obtained through direct sequencing of microbes from these habitats?; and 4) within
habitats, do container wall-associated microbial communities reflect the communities
observed on leaves and in the water column?

Materials and Methods

43

Field. Two locations with tree holes containing populations of 0c. triseriatus were used
for all field studies. The ysites, Tourney and Hudson woodlots, are located on the
Michigan State University campus (East Lansing, MI) and have been utilized for
previous studies of 0c. triseriatus (Kaufman et al. 2001, 2008). Tree holes in these
woodlots are primarily associated with American beech F agus grandifolia Ehrh.

An array of 18 tires was placed in the woodlots in the fall prior to the study year
and filled with locally-collected rainwater and senescent leaves. Also added was a
composite inoculum of microorganisms, consisting of water and particulates obtained
from nearby tires used in previous years and known to harbor 0c. triseriatus. In addition,
12 tree holes were randomly selected from the woodlots for comparison with tire habitats.
Upon hatching of natal populations of 0c. triseriatus larvae in tree holes the following
spring (mid-April), we tethered two oven-dried senescent oak leaves obtained from
Kellogg Biological Forest (Augusta, MI) into tires and tree holes using fishing line. Six 2
cm2 tiles simulating container wall material, consisting of rubber drive belt material or
tree bark, were also placed into tires and tree holes, respectively, for assessments of the
container wall microbiota. Tree bark (beech tree) was obtained from tree fall within the
woodlots and cut into tiles with a standard band saw. Leaf and container wall substrates
were allowed to condition with microbiota for three days prior to the onset of the
experiment on day 0. After conditioning, the tire array was seeded with newly-hatched
first instar Oc. triseriatus taken from our laboratory colony, propagated the previous
summer fiom the woodlots described herein. Larvae were added at densities of 0, 60, or
300 per tire, roughly approximating densities of 0, 40, and 80 larvae per liter. Each

mosquito density was represented six times per treatment. The density of native 0c.

44

triseriatus populations in tree holes was estimated on day 0 by subsampling larvae with a
syringe for counting in an enamel pan. In each container type, this procedure was
repeated biweekly to assess mosquito densities and larval development until the majority
of the initial hatch has emerged as adults. All larvae were returned to their respective
containers after being counted and scored by instar.

Leaf, container wall, and water column samples were taken on day 15 and day 30
for analysis of microbial community structure. Leaf samples were procured aseptically
using an 11 mm diameter cork borer. Additional water column and leaf samples were
taken also on day 0 to establish baseline microbial communities prior to larval feeding.
These dates were selected for microbial community sampling according to the
development time of 0c. triseriatus, which typically undergo pupation two weeks after
hatching. Leaf and container wall samples were placed in sterile phosphate buffer upon
collection, and then sonicated for 12 min. on ice to remove loosely attached
microorganisms. This method has been used in previous tree hole studies to obtain the
fraction of microorganisms available to foraging mosquito larvae (Kaufman et al. 2008).
DNA from samples was extracted for use in t-RFLP analysis and sequencing, as
described below.

Laboratory. Microcosms were constructed concurrently with the field experiment to
evaluate microbial communities at constant mosquito densities. Two independent
experiments, each consisting of six replicates, were conducted to evaluate tire and tree
hole container wall microbial communities. For each experiment, we stocked microcosms
with 500 ml deionized water and a l g senescent red oak leaf pack obtained fi'om Kellogg

Forest. This method of microcosm construction has been described for previous studies

45

of tree hole dynamics (Kaufman et al. 2001, 2002, 2006). Two container wall tiles
(described above), consisting of either tree bark or tire rubber, were added to tree hole
and tire-simulating microcosms, respectively. Finally, each microcosm received a
microbial inoculum (3 ml) of field-collected water and particulates obtained from a
composite of tree hole or tire habitats. After a three day incubation period microcosms
received either 0 or 40 newly-hatched first instar 0c. triseriatus larvae. Water column,
leaf, and container wall samples for microbial community analysis were taken, as
described above, on days 15 and 30. Additional water column and leaf samples were
taken on day 0 prior to the addition of larvae to assess the affects of larval feeding on
microbial communities.

Molecular analysis. From each experimental array (field tires and trees and laboratory
microcosms simulating the same), we created composite samples for each treatment.
Composites samples consisted of ca. 4 ng of DNA from individual treatment replicates.
Separate composites were made for leaf, water, and container wall samples. Bacterial
rRNA gene sequences were obtained using primers (63F: 5’-CAG GCC TAA CAC ATG
CAA GTC-3’ and 1387R: 5’-GGG CGG WGT GTA CAA GGC-3’) targeting a 1,300 bp
consensus region (Marchesi et al. 1998). PCR reactions consisted of ca. 10 ng composite
DNA, 4 ul each of forward and reverse primers, 50 ul F ailsafeTM PCR PreMix buffer B
(Epicentre, Madison, WI), and 1 ul FailsafeTM PCR enzyme (T aq). PCR-grade water was
added to bring the reaction to a final volume of 100 pl. Reactions were subjected through
one cycle of 94 °C for 2 min, followed by 30 cycles of 94 °C for 45 s, 68 °C for 305, and
72 °C for 1 min 30 s and a final cycle of 72°C for 7 min. Fungal rRNA gene sequences

were amplified with the primers nu-SSU-0817-59F (5’-TTAGCATGGATAATR

46

RAATAGGA-3 ’) and nu-SSU-1536-39R (5’-TTGCAATG CYCTATCCCCA-3’).
Amplification of fungal sequences was canied out as described above for bacterial
samples, except the annealing temperature was lowered to 58 °C, and extension at 72 °C
was reduced to 1 min per cycle. Amplicons were purified using the QIAquick® PCR
purification kit (Qiagen, Valencia, CA) and quantified using the Quant-iT
spectrophotometric assay (Invitrogen, Carlsbad, CA).

Bacterial and Fungal rRNA gene sequences. For field and laboratory
experiments, bacterial and fungal clone libraries were constructed by sequencing
approximately 94 randomly chosen clones per mosquito density and container substrate.
For economic and practical purposes, composite samples were created from individual
treatment replicates. For field and laboratory tire and tree hole substrates, 16S and 18S
rDNA was amplified as described above. In the case of tire microcosms, sufficient DNA
was not available for the creation of 18S rRN A gene sequence libraries following T-
RFLPs; therefore, these samples do not appear in subsequent analyses.

Clones were obtained by ligating the template DNA into pGEM-T easy vectors
and transforming the vectors into competent Escherichia coli J M109 cells (Promega,
Madison, WI). After 24 h, transformants were screened by plating on S-Gal/ampicillin
(100 ug/ml) agar (Sigma, St. Louis, MO) and randomly selecting white colonies for
isolated, overnight incubation at 37°C in ampicillin-treated (100 ug/ml) Luria-Bertani
media. Plasmids were purified (and sequenced at the Michigan State University Research
Technology Support Facility (RTSF).

Nonchimeric 16S and 18S rRNA gene sequences were classified using,

respectively, the Ribosomal Database Project classifier program

47

(http://www.rdp.cme.msu.edu, release 9.59) and the Basic Local Alignment Search Tool
(BLAST) available from the National Center for Biotechnology Information website
(http://wwwncbi.nlm.nih.gov).

T-RFLP analysis. The PCR-based T-RFLP method was used to examine
community shifts among bacterial 16S rRNA and fungal 18S rRNA gene abundances.
Shifts in the abundance of particular sequences are indicated by differences in the relative
peak areas in the T-RFLP profiles obtained for each treatment replicate as the same
concentration of PCR product was digested for each reaction. For bacterial and fungal
genes, triplicate 100 pl PCR reactions were amplified under the conditions described
above, except the forward primers for each gene region were fluorescently labeled with
FAM (carboxyfluorescein) for detection by capillary electrophoresis (IDT). For each
sample, amplification products from the three reactions were pooled during purification
then digested (ca. 10 ng) with 20 U of MspI (New England Biolabs, Cambridge, MA) in a
20 ul reactions for 3 h at 37°C. Reactions were stopped with a 20 min incubation at 65°C,
followed by an ethanol precipitation to remove excess salts fi'om the reaction prior to
capillary electrophoresis. Samples were submitted to the Michigan State University
research technology support facility (RTSF) for fragment identification with a PRISM
3100 Genetic Analyzer (Applied Biosystems, Foster City, CA) at). Peak sizes and
integrated areas under T-RFLP peaks were determined using Genescan Analysis software
(version 3.7, Applied Biosystems). Binning of T -RFLP fragments was conducted using
the T-RFLP Stats tools (Abdo et al. 2006) and R (R Development Core Team 2004).
Statistical analyses. Microbial communities present in field containers and laboratory

microcosms were analyzed using principal component analysis (PCA) of the transformed

48

(logratio) percentage of peak area represented by T-RFLP fragments (J MP® Statistical
Discovery Software, V5.1 (SAS Institute, Cary, NC.). The first two principal
components (PCs) from field tires and microcosms community were used as variables for
analysis of variance (AN OVA) to explore the effects of mosquito density and time on the
structure of bacterial and fungal communities. Because mosquito density was a
continuous variable in tree holes, PCs obtained from tree hole TRFLP fragments were
subjected to logistic regression to determine whether microbial community changes
correlated with mosquito density (Proc Reg, SAS V9.1, SAS Institute).
Results
Tree holes
T -RFLPs. Within the water column, significant changes in the bacterial

community occurred over time (PC2), but not on leaf or container wall substrates (Table
3.1, Figure 3.1). Samples taken on day 0 exhibited significant shifts in composition by
day 30 (Tukey’s HSD: p = 0.04). In contrast, significant changes in fungal communities
associated with leaves, container wall, and water column were evident over time (Table
3.1, Figure 3.3). Fungal communities changed significantly in response to time along PCl
(all substrates) and along PC 2 (water column). Shifts in fungal composition occurred
between day 0 and day 15 (Tukey’s HSD: water column, p= 0.0005; leaf surface,

=0.01), and between day 15 and 30 (p=0.002). On day 15 of the experiment, there were
no significant correlations between mosquito density and PC] scores from leaf, tile, or in
water column samples (p > 0.05; R2 = 0.32, 0.12, 0.27, respectively).

Clone libraries. For each composite tree hole sample, an average of 83 16S rRNA

gene sequences (over 1000 total) were used for classification. The distribution of class

49

level bacterial taxa obtained fiom 16S rRNA gene libraries appeared to differ among the
substrates sampled from field tree hole communities on day 15. Leaf and container wall
samples were dominated by high percentages of sequences from Alphaproteobacteria,
Garnmaproteobacteria, and Bacteroidetes (Figure 3.2). Water column samples were
dominated by the former two classes, although members of Bacteroidetes were noticeably
absent from most water samples. Mosquito density appeared to have the largest impact
on leaf-associated group, as indicated by an increase in the percentage of Bacteriodetes
with mosquitoes and concurrent reductions in the percentage of Alphaproteobacteria.
An average of 81 sequences from tree hole composites from day 15 (over 1400
total) were used for 18S rRNA gene classification. Fungal communities on leaves and in
water were characterized by a lower richness compared with container wall samples,
though all substrates were generally dominated by the presence of Sodariomycetes
(Figure 3.4). Water column fungal communities were not evaluated in response to low
mosquito densities in this experiment, as insufficient template DNA was available for
amplification; however compared to the medium larval density treatment, water samples
in high larval densities treatments had a relatively greater abundance of Dothideomycete
et Chaetothyriomycetes. A similar trend occurred in samples taken from container wall
substrates. In addition, Leotiomycetes on both leaf and container wall surfaces also

declined with increasing mosquito density.

50

Table 3.1. Analysis of variance results (ANOVA) for principle component (PC) values
obtained from relative abundance of T-RF LP fragments in bacterial (l6S rRNA) and
fungal (18S rRNA) communities in tree holes. Shown are F values (F), degrees of

freedom (df) and p values (P) for the main effect of sampling time for each substrate.

 

 

 

 

PCl PC2

Substrate F df P F df P
Bacteria

Leaf 1.19 2,24 0.32 0.61 2,24 0.55

Water 2.71 2,24 0.09 4.76 2,24 0.02

Tile 0.21 1,16 0.65 0.76 1,16 0.4
Fungi

Leaf 5.3 2,21 0.01 0.86 2,21 0.44

Water 5.23 2,22 0.01 10.31 2,22 0.0007

Tile 14.69 1,16 0.002 0.5 1,16 0.49

 

51

 

 

 

 

 

 

 

 

 

 

 

 

 

2: A “1 0 B
‘5‘3 A l 5i
I; . 41 CD
* O 1 O
.. 2- I O
”'2 A O A 9. ~
0‘: I O 0‘ I I I g
O O ‘0 '2‘. A: .
-I-r IA 4‘5
4.5% O O .35
I E I
-2 'UU'IIIWTTU7‘11IVYIII'I'VIII'IIIIIIIU' In I'WT'I‘TrTI'VIII'IIITVITIIIIIIITI‘U'I‘Y
3 -2.5 -2 4.5 -r as o 0.5 r 1.5 -12 40 -o -s 4 -2 o 2 4 s
n.
s. C PC1
5-3 I
4-3
3-2 III A Day0
2-3
15 I I Day15
3. I
0i D
: a 30
"E O . 8 O y
-23 O I
. I
: 00
.3.§ O
-‘ I'III'V"[TTTTTTT'I‘IV'TYIYYIV

 

 

 

4 -3 -2 -1 ll 1 2

P01

Figure 3.1. Principal component analysis of bacterial l6S rRNA gene communities from
tree hole habitats. Panel A: Leaf, B: Water, C: Tile. PC axes l and 2 explained 24, 17 ,

and 26% of the variation for the respective substrates.

52

100

60

40

20

100

Percent sequences

80

60

40

20

 

 

 

 

day 15

 

 

 

 

 

 

 

 

 

day 0
4 Low
day 0 day 15
Low Low Medium High

0 Alphaproteobacteria

Gammaproteobacteria
I Bacteroidetes (class)
I Other

 

100 day 15 C

 

 

 

Larval Density

Figure 3.2. Percentage of bacterial 16S rRNA gene sequences in the class taxonomic

level for composite samples taken fi'om substrates in tree holes. A Leaf, B Water, C Tile

(n = 4 per composite).

53

 

 

 

 

 

 

 

 

 

 

 

 

: T
1.5-: ‘A
1%
0.5-f I. I
E '0 0 c9 0
D1 0
g o
43.51 0 8
-I-II I
3 I
4.51
-2.UU' II'Yl'IT'IVV—UVIUUVTIY—TFTTTT‘UTI
4.5 -o.5 o 0.5 1 1.5 2
2 PC1
3 I C
1.5-: I
"i A Day0
- O
0.5-1 00 0
o-E o I Day15
: O - I
as: I O Day30
. °-
4.}
E I
4.5—
3 I

 

 

 

Figure 3.3. Principal component analysis of ftmgal 18S rRNA gene communities from
tree hole habitats. Panel A: Leaf, B: Water, C: Tile. PC axes 1 and 2 explained 20, 19,

and 28% of the variation for the respective substrates.

54

Percent sequences

 

 

  
  
 

I Low
Medium

1] High

  

 

 

 

 

 

 

 

Figure 3.4. Percentage of fungal ISS rRNA gene sequences in the class taxonomic level
(or above) for composite samples taken from substrates in tree hole habitats on day 15. A

Leaf, B Water, C Tile (n = 4 per composite). Inset legend refers to mosquito density.

55

Tree hole microcosms

T -RFLPs. Unlike field tree holes, bacterial communities in the microcosms
exhibited significant shifts over time on leaf and container wall surfaces in addition to
significant shifts in the water column (Table 3.2, Figure 3.5). For each substrate, PC1 was
associated with community changes in response to time. Samples taken at 30 days were
significantly different from samples taken at the onset of the experiment (Tukey’s HSD:
leaf, p= 0.0006; water, p<0.0001; wall, p<0.0001). In addition, mosquito presence also
had a significant effect on the structure of bacterial communities on leaves (PC1) and in
the water column (PC2).

Clone libraries. An average of 92 sequences from tree hole microcosm samples
was used for 16S rRNA gene classification (over 1300 total). Although no effect of
mosquito presence was evident in the distribution of bacterial taxa across the substrates, a
shift in dominance was apparent from day 0 to day 30 in leaf and water samples (Figure
3.6). The relative abundance of Alphaproteobacteria increased from a range of 47-57% of
total sequences on day 0 to 69-90% by day 30. Concurrently, the relative percentage of
Gammaproteobacteria declined from 22—40% of total sequences on day 0 to 0-15% on
day 30.

For 188 rRNA gene classification, we used an average of 73 sequences from tree
hole microcosm samples (over 1000 total). The abundance and richness of fungal taxa
from microcosms simulating tree holes was relatively greater in samples taken from leaf
surfaces compared with those taken from container wall or water column (Figure 3.7). As
observed for field tree holes, Sodariomycetes were a dominant member of the fungal

community from each of the substrates analyzed, although the relative abundance of this

56

group fluctuated differentially among the substrates with mosquito density. The relative

abundance of mitosporic Ascomycota tended to be greater in leaf, water, and container

wall samples from microcosms with no larvae, despite an absence of any temporal trend

for these treatments.

Table 3.2. Analysis of variance results (ANOVA) for principle component (PC) values

obtained from relative abundance of 165 rRNA gene T-RFLP fragments in tree hole

microcosm communities and showing F values (F), degrees of freedom (tit) and p values

(P) for the main effects of sampling time and mosquito density on each substrate.

 

 

 

 

PC 1 PC2

Substrate df F P F P
Leaf

Date 2,30 10.3 0.0004 1.21 0.31

Density 1,30 6.91 0.01 1.83 0.19

Date*Density 2,30 0.68 0.51 4.55 0.02
Water

Date 1,20 24.05 <0.0001 2.41 0.14

Density 1,20 1.39 0.25 6.13 0.02

Date*Density 1,20 0.75 0.4 6.24 0.02
Tile

Date 1,20 36.7 <0.0001 0.58 0.45

Density 1,20 0.04 0.85 0.04 0.84

Date*Density 1,20 0.73 0.4 0.02 0.9

 

57

 

 

11.15 11.15 .

 

 

 

 

 

 

 

 

 

 

 

 

 

A e
I : A
11.1 I 11.1-
: A ‘ ‘
: A
11115 “ 5:) 11.1151 ‘
- 00 I A; . .
9‘ n‘ 3 'll. '.
: I
41.115 ‘ I 41.1151 ‘ II
‘ d 0‘ 1 A
-0.1 . 43.1 1: A
I A
0015' I 1" l I'U'U' 0.15 I'U'lIUIUII'I'I'V'UIIUVUIIUII
0 41.15 41 41.115 11 11 115 11.1 41.15 41.1 41.115 11 11.115 11.1 11.15
PC1
11.15 ‘
3 C
11.1: I
1
11.115{ I . I 00 A Day 0
I I O O
11- 0 Ba 15
. . o 803 I Y
41115-1 I l 0 Day 30
41.1 —' I I
-ol15- 'II’I’I'UIU"U‘U'TTfiTfiU'IUVUVIUrUT
-o.2 41.15 41.1 41.05 11 11.115 11.1 11.15
PC1

Figure 3.5. Principal component analysis of bacterial 16S rRNA gene communities from
tree hole microcosms. Panel A: Leaf, B: Water, C: Tile. PC axes 1 and 2 explained 45,

50, and 69% of the variation for the respective substrates.

58

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

100 day 0 day 15 day 30 A
80 ~ .
Cl Alphaproteobactena
60 F T Gammaproteobacteria
g 40 l I BacterOIdetes (class)
5 I Other
3 20
E O 40 0 0 40
8 100 dayo day 15 day30 B 100 day 15 day30 C
.__ _
111
r
60 1 60 ' —1
40 l 40
2° 20
o J 1 .. o

 

0 4O 0 40 0 40

No. Larvae

Figure 3.6. Percentage of bacterial 16S rRNA gene sequences in the class taxonomic
level for composite samples taken from substrates in microcosms simulating tree holes. A

Leaf, B Water, C Tile (n = 6 per composite).

59

Figure 3.7. Percentage of fungal 18$ rRNA gene sequences in the class taxonomic level
(or above) for composite samples taken from substrates microcosms simulating tree holes
on days 0, 15, and 30. A, B Leaf, C, D Water, E,F Tile (n = 6 per composite). Panels A,

C, and E are microcosms with O larvae, B, D, and F are microcosms with 40 larvae.

60

 

Itufim
p

 

&

x: Z}
ff

0*
.19
0.4“
f

1, «1::
M5. ‘

#939 6?
<19
5&3?
of“? 196‘ 3:”? Q $99
W5 .

4’

&

if“
fife? «A?

 

 

Percent sequences

 

 

 

 

 

 

 

4’

é

' 1 l 1 1 1 i 1 r
-3. ”1". .‘_ . -.-. 1, . F
._ ”‘Ir -. 1
:-‘ x9: 3‘.’ - 3’.
_. .-_.- F
1.15: r‘
.4‘
a. a l
a: 3., '
5.4' g ’
1
4
. , ‘i
1 i .
1
. ..

 

 

 

 

 

U

 

 

 

 

 

lllll

 

 

61

 

 

Tires

T -RFLPs. Along PC1, bacterial communities fluctuated, reflecting significant
changes between day 15 and day 30 on leaves, container walls, and in the water column
(Tukey’s HSD: p=0.05, <0.0001, and0.01, respectively). Leaf surface communities also
shifted over time along PC2, between day 0 and day 15 (p=0.01) and between day 15 and
day 30 (p=0.02) (Table 3, Figure 8).

Fluxes in the fungal community structure were evident on leaves along PC2, with
significant shifts occurring from day 0 to day 15 (p<0.0001). Container wall surfaces did
not reflect temporal variation in the fungal taxa present (Table 3.4, Figure 3.10). As in
tree holes, correlations between mosquito density and PC1 scores from leaf, water, and
tile samples were not significant (p > 0.05; R2: 0.03, 0.01, 0.26, respectively).

Clone libraries. An average of 87 sequences from composite tire samples was
used for 16S rRNA gene classification. In comparison to tree hole communities, the
overall diversity of class level bacterial sequences appears less diverse across the
substrates sampled from tire communities (Table 3.4, Figure 3.9). Alphaproteobacteria
and Betaproteobacteria were conspicuously absent from field tire communities, although
Flavobacteria were apparent. Other major bacterial groups seen in tire samples were
Gammaproteobacteria and Bacteroidetes.

An average of 69 sequences from composite tire samples was used for 188 rRNA
gene classifications. Fungal communities were generally dominated by Sodariomycetes
across each substrate, although the relative abundance of this group varied with mosquito
density within substrates (Fig 3.11). Tremellomycetes and Cryptomonads appeared in

water samples, but comprised 57% of fungi associated with leaf surfaces or container

62

walls. In contrast to the other tire substrates, fungal communities associated with

container walls were dominated by unclassified fungi. Finally, in contrast to tree hole

communities (field and microcosm), Chytridomycota was not represented in the tire

fiingal communities.

Table 3.3. Analysis of variance results (AN OVA) for principle component (PC) values

obtained from relative abundance of T-RFLP fragments in bacterial (l6S rRNA)

communities in tires. Shown are P values (F), degrees of freedom ((11) and p values (P)

for the main effect of sampling time and mosquito density for each substrate.

 

 

 

 

PC1 PC2

Substrate df F P F P
Leaf

Date 2,28 3.09 0.06 13.26 <0.0001

Density 2,28 0.96 0.40 0.93 0.41

Date*Density 4,28 1.61 0.20 0.54 0.71
Water

Date 2,36 4.15 0.02 2.37 0.11

Density 2,36 6.07 0.005 0.37 0.69

Date*Density 4,36 0.58 0.68 0.26 0.90
Tile

Date 2,24 111.3 <0.0001 0.01 0.90

Density 1,24 0.75 0.48 1.80 0.19

Date*Density 2,24 2.58 0.10 1.77 0.19

 

63

Table 3.4. Analysis of variance results (ANOVA) for principle component (PC) values
obtained from relative abundance of T-RFLP fragments in fungal (18S rRNA)
communities in tires. Shown are F values (F), degrees of freedom ((11) and p values (P)

for the main effect of sampling time and mosquito density for each substrate.

 

 

 

 

PC1 PC2

Substrate df F P F P
Leaf

Date 2,44 229.32 <0.0001 0.20 0.82

Density 2,44 1.11 0.34 0.96 0.39

Date*Density 4,44 0.89 0.47 0.76 0.56
Water

Date 2,28 3.13 0.06 2.76 0.08

Density 2,28 3.07 0.06 0.35 0.70

Date*Density 4,28 0.03 0.97 0.68 0.51
Tile

Date 2,27 1.11 0.30 2.10 0.16

Density 1,27 3 .08 0.06 0.79 0.46

Date*Density 2,27 0.00 0.99 2.39 0.1 1

 

64

PCZ

 

 

 

 

 

 

 

 

1.5 '-—'
3 I

‘1

j I
11.55 ‘ Iii '

l A ‘ .

0" “-A I -

a , 1 _
41.5‘ A o

3 A o

.1

~1-j O o

. o A

i n A
-1I5 'I’Y'l"I'WVYIIIIYYI'Y‘YIIYYII'U'T
-1.5 -1 41.5 11 11.5 1 1.5 2
5.

5—f I

4-3

:1

3%

3 '0

21

3 O

1.:

: 0
1H I 083

a if. 09c
"1 I o

: . O
-3‘UIUU[WYY'UI'TIUU‘UIYI'IIerj
-4 -:1 -2 -1 11 1 z

PC1

 

>

N

 

 

 

 

 

 

o. B
1.5-5 0
“I o 0 o
: d) ‘ 0‘ i
0.51. A ‘
E o o 0 il
01 I
1 ‘ A
415-} .¢‘ ‘
1 O
. A
4'. (th
4.5% A
-21VfiIIIUIIII’IIIIYIVFITIIll![VIIIIII'
4.5 -l -0.5 0 0.5 I 1.5 2
PC1
A Day0
I Day15
Q Day30

 

 

Figure 3.8. Principal component analysis of bacterial l6S rRNA gene communities from

tire habitats. Panel A: Leaf, B: Water, C: Tile. PC axes l and 2 explained 20, 19, and

28% of the variation for the respective substrates.

65

100

 

 

 

 

 

 

 

 

 

 

 

 

 

day 0 day 15 A
30 . I Gammaproteobacteria
D Bacteroidetes (class)
60 1 I Flavobacteria
'3 Other
40 .
m 20
§
5 ° _
g Low Low Medium High
‘E 100 100
g day 0 day 15 3 day 15 C
111
CL 80 110 ~
60 a) .
4O 4O 1
20 20
0 o A ‘T 1
Low Medium Low Medium High

Larval Density

Figure 3.9. Percentage of bacterial 16S rRNA gene sequences in the class taxonomic
level for composite samples taken from substrates in tire habitats. A Leaf, B Water, C

Tile (n = 4 per composite).

66

 

 

 

 

WIUIUUIIIVI’Y

PC2

 

 

 

 

PC1

Figure 3.10. Principal component analysis of fungal 18S rRNA gene communities from

tire habitats. Panel A: Leaf, B: Water, C: Tile. PC axes 1 and 2 explained 24, 23, and

14% of the variation for the respective substrates.

67

 

 

PC1

 

 

0.}

Day 0
Day15
Day 30

 

 

 

100

 

A
Low
80 . I

60 [:3 Medium
40 [I High

20

 

60

 

40

Percent sequences

20

 

 

 

 

Figure 3.11. Percentage of fungal 18S rRNA gene sequences in the class taxonomic level
(or above) for composite samples taken from substrates in tire habitats on day 15. A Leaf,

B Water, C Tile (n = 4 per composite). Inset legend refers to mosquito density.

68

Tire microcosms

T-RFLPs. The presence of mosquitoes in microcosms significantly impacted bacterial
communities in the water column along both principle component axes (Table 3.5, Figure
3.12). In addition, PC2 corresponded to temporal changes in water column bacterial
communities, with significant shifts occurring between samples taken at the onset of the
experiment and those taken on day 15 (Tukey’s HSD: p = 0.01) and day 30 (Tukey’s
HSD: p < 0.0001). A significant change in the community also occurred from day 15 to
day 30 (Tukey’s HSD: p<0.0001). Bacterial communities associated with tiles simulating
tire container walls were significantly affected by the presence of mosquitoes along PC;
however, sampling date did not impart significant changes on tire communities along
either axis.

Clone libraries. As in the field experiment, the bacterial sequences present in
microcosms simulating tire habitats appeared to be less diverse than in tree hole
microcosms (Figure 3.13). In contrast to field tire communities, however,
Alphaproteobacteria were present in the tire microcosms despite exhibiting decreased
abundance over time. In general, Alphaproteobacteria and Gammaproteobacteria

dominated tire microcosms over time for each substrate.

69

Table 3.5. Analysis of variance results (ANOVA) for principle component (PC) values

obtained from relative abundance of 16s rRNA gene T-RF LP fragments in tire

microcosm communities and showing F values (F), degrees of freedom (111) and p values

(P) for the main effects of sampling time and mosquito density on each substrate.

 

 

 

 

PC1 PC2

Substrate df F P F P
Leaf

Date 2,28 1.81 0.23 1.64 0.21

Density 1,28 0.47 0.49 0.26 0.62

Date*Density 2,28 1.36 0.27 0.5 0.61
Water

Date 2,45 3.78 0.03 52.7 <0.0001

Density 1,45 52.1 <0.0001 10.7 0.002

Date*Density 2,45 1.24 0.3 0.83 0.44
Tile

Date 1,16 0.11 0.74 0.15 0.71

Density 1,16 6.14 0.02 0.54 0.47

Date*Density 1,16 0.00 0.97 0.19 0.67

 

70

d
N

 

d
O N I a I a
AllnLAlLAAJJJALAAAIAAIIAAA

I
N
1

 

.‘1

‘ A

A
0 I
‘03 0 ‘$OA Q
I o
‘ I

O

 

 

PC2

-2.5

V‘Tm'V'IIIIl'l'UU'UIVI'UIYUIVVYIIIVTT

0.5 1

-2 -1 .5

-1

-0.5 II

1.5

 

 

4.5 d

O

I
I
000

 

 

-1.

Figure 3.12. Principal component analysis of bacterial 16S rRNA gene communities

from tire microcosms. Panel A: Leaf, B: Water, C: Tile. PC axes 1 and 2 explained 29,

'UUIIY'UTIU

5 -1 41.5

IFT'U'Y'IUYUU'U

0 0.5 1

PC1

UUIIII'I

1.5 2

-A
U'l

d

.51
1:1

I
d

1
d
U!

1
N

 

 

 

 

 

.3 “I
: g4
.: A A
I A‘ '- At.
1 I. l
3 I
‘3 O ' 036,
~ 0
.- o
a oCi °
3 0 o
qIIIYIUIIU'IUIIIIVIIIUIUIIIVUIIIYU'TT
-2 4.5 -1 41.5 0 0.5 I 1.5 2
PC1

A DayO

I Day15

O Day30

 

 

 

42, and 58% of the variation for the respective substrates.

71

 

 

 

 

 

 

 

 

 

 

 

   

da 0 a. 15 da 30 A
80 y _ y y
60 : D Alphaproteobacteria
‘ i I Gammaproteobacteria

40 “ I Other
03
8 3
c 20
° 1
3
o-
3 0
.. 0 4o 0 4o 0 40
E
g dayo‘ ' day '15 * ' ”da‘y30 B day1'5 ’day30 ” C
a) 80 . so.
0.

60 ‘ } , _: 60

40 ‘ , 1 4o

‘ l
20 i h ‘ , 20
o .4 ,1. 0
o 40 o 40 o
No. Larvae

Figure 3.13. Percentage of bacterial 16S rRNA gene sequences in the class taxonomic
level for composite samples taken from subsnates in microcosms simulating tire habitats.

A Leaf, B Water, C Tile (n = 6 per composite).

72

Discussion

To our knowledge, this is the first study to address the composition of microbial
communities associated with tire container habitats. Evaluating the variation in tree hole
versus tire habitats is an important consideration, as mosquito species often segregate on
the basis of container type. The differential capacity among mosquitoes for vectoring
diseases may be relevant, as competitive displacement among species may occur in
specific habitats, resulting in a superior disease vector supplanting a mosquito species
that is more innocuous fi'om a public health perspective. For example, the invasive
mosquito 0c. japonicus, which has spread through much of the Eastern United States in
recent years (Andreadis 2001), has recently begun to encroach upon tire habitats in
Michigan (personal observation). If time proves this species to be a superior competitor,
native populations of 0c. triseriatus may be supplanted.

Temporal fluctuation in the structure of bacterial and fungal communities was the
most evident effect observed in field and microcosm experiments for both container
types. Bacterial communities obtained from leaf, water, and container wall samples
shifted significantly from day 0 to day 30, with only one exception. Changes in leaf-
associated bacterial communities were evident in tree hole microcosms, but not in field
tree holes. Fungal communities also fluctuated among all three substrates in field tree
holes; however, among tire substrates, only shifts in leaf-associated communities were
evident. Temporal shifts in community structure are possibly related to changes in the
surrounding environment, including disturbances in the habitats related to temperature,
precipitation, and wind-home introduction of new microorganisms to the communities.

These factors are less likely to affect temporal changes in microcosms, as such

73

environmental variables are carefully controlled. Instead, it is possible that bottom-up
effects, such as the build-up of waste products from developing larvae or by-products
associated with leaf decomposition, may also be contributing to shifts in microorganism
community structure over time.

Contrary to what we might have expected based on previous studies, larval
density generally did not affect the communities of bacteria or fungi represented by T-
RF LP fragments from field tree holes or tires. Indeed, significant larval affects on
principal component variables representing bacterial communities were only observed in
the water column of the latter. Mosquito density did influence bacterial communities in
tree hole and tire microcosms, however. Effects in these habitats manifested in the water
column and, for tree hole microcosms, in association with leaf material. Several factors
may have contributed to the absence of mosquito effects in field sample, including
overall variation in the field experiments. Variation not due to mosquitoes, including
stemflow and the presence of alternative leaf and other detrital material, may have
obscured the effect of larval feeding on microbial communities.

The communities of heterotrophic bacteria observed in this study were generally
dominated by Alphaproteobacteria and Bacteroidetes, groups commonly associated with
aquatic habitats (Kirchman 2002). As described by Kaufman (2008), leaf-associated
Alphaproteobacteria in tree holes decreased with larval presence. Within this group,
Caulobacteraceae decreased with larval density. The Porphyromonadaceae within the
Bacteroidetes phyltun, also fluctuated in response to mosquito density. As in Kaufman et

al. (2008) the abundance of this group decreased with mosquito feeding pressure;

74

however, in the current study the effect was observed on container wall tiles rather than
leaf surfaces.

In contrast to the variability of bacterial groups at the class level in tree holes, all
substrates in tire communities and leaf material in tire microcosms were dominated by
Gammaproteobacteria, with little apparent response to mosquito feeding. Unlike other
Proteobacteria, Gammaproteobacteria tend to be less abundant in freshwater habitats
(Kirchman 2002). This difference suggests that the two communities may be
differentiated, from the mosquito point of view, with respect to the quality of microbial
food resources available. It has been observed in microcosm studies that some
Gammaproteobacteria, the Enterobacteriaceae, may increase in the presence of mosquito
larvae, and that this group may be more resistant to digestion given its regular
colonization of animal guts (Kaufman et al. 1999). Betaproteobacteria were absent from
each of the habitats samples, which is unexpected given the preponderance of this group
in aquatic habitats and previous tree hole studies (Kirchman 2002, Kaufman 2008).

The relatively high frequency of 18S rRNA gene sequence in tire habitats
classified only to the level of Fungi suggests a preponderance of unique species
compared with tree hole habitats. Sodariomycetes were highly represented among each
container type. Within the Sodariomycetes, Diaporthales was the most represented order,
a fact that is unsurprising due to the utilization of freshwater habitats by members of this
group (Samuels and Blackwell 2001; Zhang and Sung 2008). Breakdown of plant
detritus, particularly lignin, is a function ascribed to several members of this group;
hence, their abundance in detritus-based habitats such as tires and tree holes is consistent

with the function expected of microbial community members.

75

The cumulative variation explained by principle components 1-3 was much lower
in our field community samples compared with those taken from microcosms. More
variation is expected in natural systems than in controlled laboratory experiments;
however, another factor that may have contributed to this difference is the addition of a
single microbial inoculum in microcosms at the onset of the experiment. In contrast to
field tree holes, new allochthonous sources of microbial groups were generally not
available from the surrounding environment. Although large introductions were
precluded by screen enclosures, it is assured that new microbial introductions into field
containers (e. g. windbome or soil-associated) contributed to a higher overall variability in
these systems that was independent of the imposed factors.

Although temporal flux was highly evident in fungal communities on tree hole
container surfaces, the container walls of tire habitats did not exhibit similar community
dynamics. The difference between materials comprising the container walls cannot be
overlooked as a critical factor that may contribute to the species-specificity of container
surfaces. Indeed, in the current study, we were unable to obtain sufficient fungal material
from tiles in microcosms representing tire container wall surfaces to proceed with fimgal
community analyses on this substrate. Bark-lined tree holes represent an additional
source of decaying organic matter for mosquitoes residing in these habitats. Although
characterized by a high C:N ratio, and thus a relatively recalcitrant source of organic
carbon compared with other forms of tree hole detritus, bark-lining in tree holes provides
an important niche for fungal grth that is absent in tire habitats. Such inherent

difference may explain the success of 0c. triseriatus in tree holes, or at least suggest a

76

possible basis for which they may be outcompeted by other mosquito species in tire
habitats.

It has become increasingly clear from recent research that fungi play a critical role
in driving 0c. triseriatus production (Kaufman et a1. 2008, Xu et al. 2008, Pelz-Stelinski
et al. unpublished). Scirtid beetle larvae, which often co-occur with 0c. triseriatus in tree
holes, are known to facilitate the production of mosquito adults (Bradshaw and Holzapfel
1992; Paradise 1999, 2000; Daugherty and Juliano 2000, 2001, 2002). Recently, we have
shown that changes in the structure of fungi on leaves and in the water column occur in
the presence of scirtid beetles (Pelz-Stelinski et al. unpublished). Any contribution to
fungal activity, be it from the presence of macroinvertebrate facilitators or increases in
available fungal niche space, should result in enhanced mosquito success. In spite of this,
we have not discovered any evidence that would indicate specific fungal groups are
consistently selected by foraging larvae. Rather, fungi appear to be a general resource
that is differentially available to mosquitoes with respect to substrate. Tire and tree hole
habitats are distinct with respect to their the structure of fungal communities, but perhaps
also in terms of the biomass they support. Although we did not quantify fimgal biomass
or activity in this study, it is apparent that fungi were less available on container walls
associated with tire habitats. Future studies should address the relative contribution of
fungal biomass to resource pools in tire habitats as a possible factor regulating mosquito

populations.

77

CHAPTER 4

MULTITROPHIC INTERACTIONS IN TREE HOLE COMMUNITIES ARE

FACILITATED BY SCIRTID BEETLES

78

Introduction

Tree holes are small, discrete ecosystems that contain heterotrophic communities
driven by allochthonous inputs of soluble and particulate organic matter. In Eastern North
American tree holes, larvae of the Eastern treehole mosquito, Ochlerotatus triseriatus
(Say), are usually the dominant macroinvertebrate consumers. Leaves are a typical source
of coarse particulate organic matter (CPOM), but animal-derived detritus, such as
invertebrate carcasses and fecal material, also supply energy to these systems (Daugherty
et al. 2000, Yee and J uliano 2006). In addition, stemflow nmoff brings dissolved organic
carbon and nutrients into the habitats (Carpenter 1982, Walker et al. 1991 , Kaufman et al.
1999, 2002). The success of 0c. triseriatus and similar mosquito species is defined by
high adult productivity and body weight at emergence, and this depends on the nutrition
they receive while in the larval stage. Microorganisms are the main nutritional resource
for mosquito larvae, which feed by browsing on container surface microbial biofilms and
filtering FPOM and microorganisms from the water column (Merritt et al. 1992). Fungi
associated with leaf detritus alone may account for around 10% of the detrital biomass in
tree hole habitats (Kaufman et al. 2002, 2006). Hence, microbial degradation of detritus
is a critical link between developing larvae and nutrients, as processing by

microorganisms incorporates nutrients and otherwise inaccessible carbon from CPOM.

In the Midwestern and Eastern United States, larval scirtid beetles (Helodes and
Prionocyphon spp.) are shredders which often co-occur with 0c. triseriatus in tree holes
(Barrera 1996; Paradise 2004). Scirtid feeding activity results in the skeletonization of
leaf detritus and the associated conversion of coarse particulate organic matter (CPOM)

to fine particulate organic matter (FPOM) in the form of small leaf particles and feces.

79

Several studies indicate that the presence of these beetles in tree holes conditionally
facilitates the survival and development of 0c. triseriatus under conditions of low leaf
litter availability by improving the quality of resources available to 0c. triseriatus
(Bradshaw and Holzapfel 1992; Paradise 1999, 2000; Daugherty and Juliano 2000, 2001,
2002). In addition to mosquito populations, processing chain benefits have been reported
to facilitate populations of the ceratopogonid midge Culicoides guttipennis in the
presence of Helodid beetles (Paradise and Dunson 1997). Similarly, processing chain
commensalisms benefiting mosquitoes have been reported in pitcher plants. By
increasing the conversion of coarse particulate plant material to particulates, the midge
Metriocnemus knabi facilitate populations of the mosquito Wyeomyia smithii in pitcher
plants (Heard 1994). Increased conversion of leaf material into FPOM is the main
mechanism cited for any benefit experienced by mosquitoes in the presence of scirtids
(Paradise 1999, 2000; Daugherty and Juliano 2002, 2003). Indeed, Daugherty and
J uliano (2003) demonstrated that additions of scirtid fecal material and its associated
microbiota to microcosms increased larval 0c. triseriatus development by providing
additional nutrient resources; however, the source of this positive effect (feces,
microbiota, or a combination) has not been determined. Furthermore, although the
conversion of CPOM into FPOM may benefit foraging larvae by being ingested directly,
it may also promote access to fungal material otherwise embedded in the leaf matrix.
Previous studies have addressed larval mosquito feeding on microorganisms in
container habitats, but microbial food resource dynamics remain ill-defined. Water-
column-associated bacteria exhibit negligible or inconsistent responses to mosquito

presence and, thus far, there have not been any published studies of fimgal communities

80

associated with tree hole water columns. Furthermore, larval feeding effects on the
composition of water column microbial communities are limited to bacteria and protists
(Kaufman et al. 1999, 2002, Cochran-Stafira and von Ende 1998, Kneitel and Miller
2002, Tryzcinski et al. 2005). In contrast, there is convincing support for the importance
of microbial communities associated with detritus in previous assessments of tree holes.
Increased fungal enzyme activity, and decreased bacterial productivity and abundance are
associated with mosquito feeding (Kaufinan et al. 1999, 2001, 2002, Kaufinan and
Walker 2006). Despite the apparent response of microorganisms to larvae, the
composition of microbial communities in tree holes has been addressed only recently
(Kaufman et al. 2008, Xu et al. 2008). In that study, larval feeding effects were evident
on the taxa comprising fungal and bacterial communities associated with leaf detritus,
with particular influence on Saccharomycetes, Dothideomycetes, and Chytridiomycota
fungal taxa, and on Alpha— and Betaproteobacteria. It is apparent that any influence of
scirtids on the microbial community may affect mosquitoes, given the direct utilization of
microorganisms by larvae for nutrient acquisition. By altering the detritus, scirtid beetles
modify the bottom-up influence of detritus on bacterial and fungal populations. Such
blending of “bottom-up” influences and “top-down” trophic cascades are described for
aquatic ecosystems, including tree hole container habitats, yet the ramifications of scirtid
beetles on microorganisms is poorly understood (Carpenter and Kitchell, 1987,1988).
Although previous studies have postulated mechanisms for facilitation of
mosquito larvae by scirtids, they are limited to measurements of macroscopic changes
such as the abundance of FPOM, population densities of mosquitoes, and decomposition

of scirtid carcasses (Paradise 1999, 2000; Daugherty and Juliano 2002, 2003). While all

81

of these factors play a role in mosquito growth, these indirect assessments lack the
intermediate microbial step linking the trophic levels. Elucidating the effect of scirtid
feeding activity on mosquito productivity requires an assessment of the changes such
feeding renders on the microbial community. In an effort to understand the ecological
influences driving mosquito production, we sought to describe microbial mechanisms
underlying facilitation of larval 0c. triseriatus grth and development by scirtid beetles
via assessments of the microbial community associated with scirtid presence in
microcosms simulating natural tree holes. We postulated that detritus-associated fungal
dynamics would be the most dramatically affected by scirtid presence as this location is
the “center of activity” for foraging scirtid larvae (Dixon and Chapman 1980).
Materials and Methods

Experimental design. The mechanism regulating the facilitation of 0c.
triseriatus larvae by scirtid beetle larvae was investigated by crossing beetle
presence/absence with mosquito presence/absence in a multi-factorial design with the
following scirtid: 0c. triseriatus ratios: 0:0, 10:0, 10:40, 0:40. Twelve replicates of each
treatment combinations were constructed, with one set of six replicates destructively
sampled midway through the experiment (ca. day 30). Because resource level mediates
the interaction between these species such that facilitation is only apparent at low
resources, low leaf litter rations were used for all replicate microcosms.

Individual microcosms simulating natural tree holes received a 1 g senescent oak
leaf pack in 500 ml deionized water. Microcosms were constructed similarly to those

described in previous tree hole studies (Walker et al. 1991, Kaufman et al. 2001, 2006). A

82

3 ml microbial inoculum, consisting of homogenized tree hole water and particulates, was
added to microcosms three days prior to the addition of macroinvertebrates.

0c. triseriatus larvae, obtained from our colony at Michigan State University,
were added as newly-hatched first instar. Scirtid beetle larvae, ranging from 2“d to 4th
instar, were obtained from local tree holes (Toumey Woodlot, E. Lansing, MI) and
replaced semi-weekly as necessary to maintain a constant population in scirtid treatments.
Dead beetles were removed to avoid confounding the effect of scirtids by providing
additional resources to mosquito or microbial populations. Due to the difficulty of
identifying live beetle larvae, scirtids were not identified to species (Paradise and Dunson
1997; Daugherty and Juliano 2001, 2003).

Sampling. Leaf condition was assessed by taking dry weight measurements of
leaf discs on day 30 and day 50. Leaf disc samples (20 mm diam.) were procured from
each microcosm on day 30 and day 50 using a cork borer. Two leaf samples apiece were
taken for assessing bacterial abundance, bacterial productivity, fungal biomass, fimgal
degradation enzymes and microbial community structure. Prior to analyses, leaf disc
samples were sonicated for 12 min in an ice bath to obtain the loosely attached, surface-
associated microorganisms, as these better represent those encountered by foraging
mosquito larvae (Kaufman 2008). In addition to leaf samples, water samples were also
taken from microcosms on each sample date. After stirring microcosms to homogenize
water and particulates, 15 ml samples were removed from microcosms for the analyses
indicated above.

Bacterial Abundance and Productivity. Bacterial abundance on the leaf surface

and in the water column was quantified from subsamples via direct microscopic counts

83

(DMCs) of bacteria using the DAPI (49,6-diamidino-2-phenylindole) fluorescent staining
procedure (Porter and F eig 1980, Walker et al. 1988). Formalin (3% formaldehyde
final concentration)-preserved samples were kept at under dark conditions at 4°C
until analysis, then filtered onto black Nucleopore filters (0.2-mm pore size; Costar,

Cambridge, Mass). After filtering, samples were stained at a final concentration of 20
11ng for 5 min. Two filters were counted (200 cells per filter) for each subsample at
1000X using a Nikon E800 fluorescent microscope (Nikon, Inc. Melville, NY).

Bacterial productivity was estimated by the incorporation of [3H]- leucine (50
Ci/mmol, Life Science, Boston, MA) into bacterial biomass using the microcentrifuge
tube method (Smith and Azam 1992, Kirchman 2001). Labeled leucine was added at a
final concentration of 25 nM. It was determined previously that saturation of uptake
kinetics occurred at 100 and 400 nM for water and leaf samples, respectively (Kaufinan
2001, 2002, 2006). Samples were incubated for 30 min at 20°C to allow for the
incorporation of labeled leucine into bacterial biomass. Following incubation, 5% TCA
(trichloroacetic acid) was added to terminate the reaction and precipitate protein. Samples
were subsequently rinsed and concentrated as described in prior microcosm studies
(Kaufman 2001, 2006). Standard liquid scintillation counting of samples was conducted
in a Beckman LS6500 scintillation counter (Beckman Coulter, Inc., Fullerton, CA).

Fungal Biomass. Fungal biomass was estimated using surrogate measurements of
ergosterol, a sterol associated with fungal cell walls (Newell and Barlocher 1993,
Suberkropp and Weyers 1996). Leaf disc subsamples sonicated as described above in
sterile microcosm water were kept in the dark and stored in high performance liquid

chromatography (HPLC) grade methanol prior to extraction and quantification of

84

ergosterol using I-IPLC and UV detection (Kaufman et al. 2001, Kaufman and Walker,
2006).

Enzyme Activity. Leaf enzyme activity was assayed through incubations of leaf
disc subsamples with two methylumbelliferyl (MUF)-labeled substrates, 4-
methylumbelliferyl-B-D-cellobioside and 4- methylumbelliferyl- B-D-xyloside, which
will respectively estimate cellobiohydrase, and xylosidase activity. These substrates are
analogous to plant polymers and thus provide an estimate of leaf-associated carbohydrase
activity (Kaufman and Walker 2006). MUF was liberated upon enzymatic cleavage of the
substrates during a 1.5 h incubation at 22°C. Unbound MUF, which fluoresces at 360 nm,
was measured using a 96-well Hoefer DyNA Quest 200 fluorometer (GE Healthcare,
Little Chalfont, Buckinghamshire, United Kingdom).

Microbial Community Analysis. The compositions of leaf- and water column-
associated bacterial and firngal communities in microcosms were assessed using terminal
restriction fragment length polymorphism (T -RFLP) analysis (Liu 1997, Marsh 1999).
Leaf disc samples were sonicated as described above in sterile phosphsate buffered saline
to dislodge surface- associated microorganisms. Thereafter, DNA was isolated fiom leaf
sonicate and water samples using the MoBio UltraClean soil extraction kit (Carlsbad,
CA). Independent amplification of DNA for T-RF LP analysis were performed with two
primers sets respectively targeting small subunit (SSU) ribosomal DNA from bacteria
(16S rDNA) and fungi (18S rDNA). A 1300 bp 168 rDNA fragment was amplified using
universal eubacterial primers 63F and 1387R (5’-CAGGCCTAACACATGCAAGTC-3’
and 5’-GGGCGGWGTGTACAAGGC-3’, respectively) (Marchesi et al. 1998). A 762

bp 188 rDNA fragment was amplified using universal fungal primers nu-SSU-0817-59F

85

and nu-SSU-1536-39R (5’-TTAGCATGGA ATAATRRAATAGGA-3’ and 5’-
ATTGCAATGCYCTATCCCCA-B’, respectively) (Bomeman and Hartin 2000).
Forward primers for each target sequence were fluorescently labeled (6-F AM) at the 5’
end (IDT Technologies). Template DNA from fimgal and bacterial communities was
amplified in three 100 11] PCR reactions per primer set using the F ailsafeTM PCR System
(Epicentre Biotechnologies, Madison, WI). Each PCR reaction consisted of ~1 0 ng
template DNA, 50 111 F ailsafeTM PCR PreMix buffer B, 1 111 F ailsafeTM enzyme, and 4 ul
of each DNA primer. Reactions were held at 94°C for 2 min, then amplified under PCR
cycle conditions optimized for each target sequence. For 16S rDNA, reaction mixtures
were cycled 30 times through the following steps: denaturing for 45 s at 94°C, annealing
for 30 s at 68°C, and extension for 1 min 30 s at 72°C. For 18S rDNA, reaction mixtures
were cycled 30 times through the following steps: denaturing for 45 s at 94°C, annealing
for 30 s at 58°C, and extension for 1 min at 72°C. Both reactions were subjected to a final
extension step for 7 min at 72°C. PCR products from triplicate reactions were co-purified
with the QiaquickTM PCR purification kit (Qiagen, Valencia, CA) and then were
subjected to independent digestions with two restriction enzymes, MspI and HhaI (New
England Biolabs, Cambridge, MA) (200 ng product with unit of Hha or Msp for 3 h at
37°C) then incubated overnight with ethanol at -80°C to remove excess salts. The size
and frequencies of each labeled fragment was determined by capillary electrophoresis of
digested samples with the PRISM 3100 Genetic Analyzer (Applied Biosystems, Foster
City, CA).

Statistical Analysis. Multivariate analysis of variance (MAN OVA) was used to

analyze related mosquito productivity variables, including total adult mass, total survival,

86

and total number of adults emerged (SAS v. 9.1, SAS Institute, Inc., Cary NC, USA).
Values were log transformed prior to analysis. Water chemistry values, bacterial
abundance, bacterial productivity, leaf enzyme activity, ergosterol content, and leaf mass
lost were analyzed using a standard analysis of variance (ANOVA ) followed by
Bonferroni correction to control the experiment-wide error (Rice 1989). In addition,
means were square-root transformed as needed to meet the assumptions of AN OVA. In
addition to the main treatment affect, the effect of experimental block on response
variables was also assessed. Upon finding that blocking was significant, individual blocks
were treated separately to determine the within-block treatment effects.

T-RFLPs were aligned and peak areas were determined using the GeneScan
software package (Applied Biosystems). A sample was discarded if its summed peak area
was less than 1000 or if a normality plot indicated the presence of significant outliers. In
addition, peaks that comprised <3% of the community or appeared in fewer than three
samples were also discarded. Common peaks were binned using the “T-RFLPs Stats”
analysis tools (iBest http://www.ibest.uidaho.edu/tools/trflp stats/index.php) REF. The
resulting matrices were converted to log ratios of the relative abundance of peak area in
each sample, then analyzed by principle component analysis (PCA) on the covariance
matrix using JMP® Statistical Discovery Software (SAS Institute, Inc., Cary NC,
USA).Scores from PC’s 1-3 and transformed percentages of TRFLP fragments were
subjected to AN OVA for treatment comparisons.

Results
Adult mass, emergence, development time, and total survival of mosquitoes did

not differ significantly in microcosms containing scirtids compared with those in which

87

scirtid beetles were absent and the lack of effect was consistent from day 30 to day 50
(Figure 4.1; MANOVA: Wilks’ Lambda = 0.78, df = 5,16, p = 0.49). In contrast, leaf
decomposition was significantly affected by the presence of sciritids in microcosms (F =
11.96; df = l, 23; p = 0.003), such that microcosms containing scirtids had a significantly
lower proportion of leaf mass remaining than non-scirtid microcosm (Figure 4.2; T ukey’s
HSD, P < 0.05). Neither mosquito presence nor the interaction between the insects
significantly affected leaf decomposition (F = 0.69; df = 1, 23; p = 0.417; and F = 0.0; df

= l, 23; p = 0.962, respectively).

88

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

10
A

:9, 8
5
g 8 6 CI Oscirtids
. 1:1
8 1T1 I10 scirtids
c E
111 ‘1’ 4
111
2

2

o

1 60
01
E - 50
7; 0.75 '1 m
w I 40 c
m 1 2..
E 5
z 30 —
3 05 ,3
111 _ 20 E
r:
111 0.25
111
E 10

o 0

Females Males Total Adults

0. triseriatus

Figure 4.1. Mosquito production variables from microcosms containing 40 0c.
triseriatus larvae. (A) Number of adults (B) Total adult mass. (C). Survival (40 initial
larvae). open characters represent microcosms with no scirtids filled characters represent

10 scirtids. Values are means :t SE; n = 12 all variables.

89

Leaf Mass Remaining

0.84 ‘ ’ ' T T71
. DOMIwMj

 

 

 

 

§
111 l
g .
2 0.76 l
“5
111
_l
C
.9
t
8 0.68
9
n.

0.60

0 1O
Scirtid Density

Figure 4.2. Proportion of leaf mass remaining in microcosms on day 50. S: Scirtid, M:

Mosquito, white bars: 0 M, grey bars : 10 M. Values are means i SE (N=24).

90

Bacterial abundance on leaf surfaces was depressed in the presence of
macroinvertebrates (Table 4.1 and Figure 4.3). Under these reduced-resource conditions,
mosquito larvae significantly lowered bacterial abundance compared with no larvae
conditions; however, scirtid presence scirtid presence significantly interacted with
mosquito presence at day 30 such that microcosms containing both insects had a greater
abundance of leaf-associated bacteria than those containing mosquito larvae alone
(Tukey’s HSD test, P < 0.05). Although scirtid beetles did not significantly decrease
bacterial abundance at day 30, by day 50 microcosms with scirtids alone exhibited lower
abundance than microcosms with no macroinvertebrates present. In the water column,
bacterial abundance was not significantly affected by the presence of either
macroinvertebrate, although a significant decrease in abundance did occur from day 30 to
day 50.

Leaf associated bacterial productivity was significantly lower in the presence of
mosquito and scirtid beetle larvae compared with the no macroinvertebrate treatment
(Table 4.1 and Figure 4.3). Productivity differences were not apparent among
microcosms receiving macroinvertebrates, likely in response to the significant interaction
between mosquito larvae and scirtid beetles. Significant reductions in water column
bacterial productivity were evident in the presence of mosquito larvae and productivity
remained unchanged whether scirtids were present or absent. In addition, water column
productivity also declined significantly from day 30 to day 50 for microcosm treatments.

The activity of leaf decomposition enzymes for the substrates xylan and
cellobiose significantly increased in microcosms containing scirtids compared with

microcosms in which scirtid beetles were absent (Table 4.2 and Figure 4.4). This effect

91

was neither significantly influenced by time nor the presence of mosquito larvae in
microcosms. In contrast, although fungal biomass (ergosterol concentration) in
microcosm leaf samples was numerically greater in microcosms containing scirtids this
effect was not statistically significant. Fungal biomass was not significantly influenced by
the presence of either 0c. triseriatus or scirtid larvae; however, the concentration of
ergosterol was significantly greater for all treatments on day 50 compared to day 30

(Table 4.2 and Figure 4.4).

92

Table 4.1. Summary of ANOVA results for bacterial abundance and bacterial

productivity values from microcosm water column and leaf material.

—
Source df F Value P Value

Water column abundance (cells/ml)

 

Mosquito 1 0. 1 5 0.699
Scirtid l 0.69 0.410
Time 1 13.88 0.001*
Mosquito*Time 1 0.00 0.975
Scirtid*Time 1 0.01 0.907
Mosquito*Scirtid 1 0.79 0.3 79
Error 37

Leaf surface abundance (cells/disc)
Mosquito 1 23.62 <0.001*
Scirtid l 0.83 0.368
Time 1 0. 5 0.483
Mosquito*Time 1 l .7 0.200
Scirtid*Time l 6.89 0.012*
Mosquito*Scirtid l 6.61 0014*
Error 40

Water column productivity (nmol leucine /ml/h
Mosquito 1 9.37 0004*
Scirtid 1 0.14 0.712
Time 1 12.91 0.001*
Mosquito*Time 1 0.8 0.378
Scirtid*Time l 0.7 0.406
Mosquito*Scirtid 1 0.25 0.617
Error 40

Leaf surface productivity (nmol leucine /disc/h)
Mosquito l 14.77 0.001*
Scirtid l 6.51 0.016
Time 1 2.35 0.135

93

Table 4.1 (cont’d).

Mosquito*Time l 2.04

Scirtid*Time l 1 .39
Mosquito*Scirtid 1 1 7.41 <0.001 *
Error 32

 

*Indicates p-values that are significant following Bonferroni adjustment

94

V.»

 

 

Dom lam

 

 

mmonmzm: 36:33:05

$083.2 38:33.2

 

ab

m3

Log Cells l ml

mb

mb

 

 

 

 

u.»

 

 

 

 

 

 

 

 

 

 

 

a

 

 

fa

. A.»

 

 

 

9m .

m3

Log Cells I disc

ah I

ma

 

 

 

mam—=6 Ab. wagon”: 339830: 85 ”8858.8. m” momma. 3” 30358. 2E8 gem” o m. mam gem” 5 m. <m_=am m8 333 .1.” mm

13.

oz—

 

 

O>< no

.52.

oz. .32

U>< Ac

 

 

 

 

 

 

 

oz.

#02.

U>< no

02.

O>< ho

ho:

1.—

 

cleH/IW/IOUJU

cOlel/3slpllowu

95

Table 4.2. Summary of ANOVA results for ergosterol and fimgal degradation enzyme

concentrations in microcosm leaf material.

Source df F Value P Value
Ergosterol (ppm / leaf disc)

 

Mosquito l 2.04 0. 1 62
Scirtid l 0.61 0.439
Time 1 12.84 0.001*
Mosquito*Time l 0.88 0.3558
Scirtid*Time 1 0.88 0.353
Mosquito*Scirtid 1 0.92 0.344
Error 36

Xylose polymer activity (ppm/disc/h)
Mosquito 1 0.40 0.533
Scirtid 1 9.96 0003*
Time 1 0.16 0.691
Mosquito*Time l 0.04 0.839
Scirtid*Time l 0.38 0.542
Mosquito*Scirtid 1 0.02 0.884
Error 40

Cellobioside polymer activity

(ppm/disc/h)
Mosquito 1 0.03 0.871
Scirtid 1 10.66 0002*
Time 1 0.48 0.494
Mosquito*Time 1 0.0 0.969
Scirtid*Time l 0.1 1 0.740
Mosquito*Scirtid 1 1 .25 0.269
Error 40

 

*Indicates p-values that are significant following Bonferroni adjustment

96

Enzyme Activity
0.5 . —

 

‘ Xylose polymer j [j 0 S I 10 s

 

 

 

0.4

0.3

 

0.2

0.1 l

 

 

 

 

 

 

 

 

ppm/disc/h

 

Cellobiose
0.3

0.2 J I

 

0.1

 

 

 

 

 

 

0 M 40 M 0 M 40 M
Day 20 Day 40

Figure 4.4. Fungal enzyme activity. S: Scirtid, M: Mosquito, white bars: 0 S, grey bars :

10 S. Values are means 3: SE (N=24).

97

Leaf Ergosterol

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

4
D 05 I 103
3
0
.2
U
1..—
(U
2 2
E
Q.
Q.
1
0
0M 40M 0M 40M
DAY20 DAY40

Figure 4.5. Fungal biomass and enzyme activity . S: Scirtid, M: Mosquito, white bars: 0

S, grey bars : 10 S. Values are means 3: SE (N=24).

Total nitrogen in microcosms was significantly greater on day 50 than on day 30
(Tukey’s HSD, P = 0.004), but was not significantly affected by either mosquito of
scirtid presence (F= 1.48; df= 1, 40;p = 0.23; and F= 1.53; df= 1, 40; p = 0.223,

respectively). Total phosphorus in microcosms was not significantly affected by the

98

presence of either macroinvertebrate in microcosm (F = 1.27; df = 4, 43; p = 0.3). In
addition, the Block effects were not significant for N (F = 0.67; df = l, 32; p = 0.41) or P
(F= 0.01; df= 1, 32;p = 0.94)

T-RFLPs

MSP. PC1 scores for bacteria in the water column were significantly affected by
time (Blocks I and II), exhibiting shifts in the community structure from day 20 to day 45
(F= 5.86, df= 1, 10, P = 0.001; F: 14.03, df= 1, 15, P = 0.002) (Figure 4.6; complete
ANOVA tables are shown in Appendix). Scirtids also had a significant effect on bacterial
communities along PC1 time in Block 11 (F= 16.93, df = 1, 15, P < 0.001, but not in
Block I. PC2 scores were not significantly affected by treatments in Block I; however, in
Block II, both time and mosquito presence significantly affected water column bacteria
along PC2 (F = 53.23, df= 1, 15, P < 0.001; F = 27.90, df= 1, 15, P < 0.001). Both
mosquito presence and time significantly affected PC1 scores for leaf surface bacterial
communities in Block 11 (F = 41.17, df= 1, 15, P < 0.001; F = 8.04, df= l, 15, P =
0.012), but the same effect was not observed in Block 1. Neither insect presence nor time
significantly affected on PC2 in either block.

Water column fungal communities exhibited significant shifts in response to
mosquito presence and time along PC2 (Block 11: F = 37.95, df = 1, 15, P < 0.001; F =
13.23, df = 1, 15, P = 0.002) (Figure 7). For both blocks, there were no significant effects
of any factor on PC1. Similarly, a shift in leaf-associated fungal communities in response
to mosquito presence was significant along PC2 in Block 11 of the experiment (F = 6.19,
df = 1, 16, P = 0.024). In general, however, no significant shifts were evident in leaf

fungal communities across the experiment in response to time or scirtid presence.

99

HHA. PC1 scores for bacteria in the water column were significantly affected by scirtids
in Block I (F = 21.06, df = 1, 11, P = 0.001). PC 2 scores representing water bacterial
communities were significantly affected by the interaction between mosquitoes and
scirtids (Block I: F = 12.94, df = 1, 11, P = 0.004) or mosquitoes alone (Block II: F =
6.19.67, df = l, 15, P = 0.01). Leaf surface bacterial communities represented by
principal component axes were significantly affected by the interaction of scirtids and
time (PC2, Block I: F = 6.63, df = l, 14, P = 0.022) or by scirtids alone (PC1, Block 11: F
= 5.09, df= 1, 15, P = 0.039).

Water column fungal communities changed in response to mosquito presence
(PC1, Block I: F = 10.02, df = l, 7, P = 0.016), and the interaction of mosquito presence
and time (PC2, Block II: F = 11.45, df = 1, 16, P = 0.004). Scirtids did not significantly
affect changes in leaf-associated fungal communities, although these communities
changed in response to the presence of mosquitoes (PC2: F = 45.52, df = l, 15, P <
0.001) and time (PC1: F: 24.29, df=l,15, P < 0.001; PC2: F: 8.65, df=1,15, P =

0.010) in Block 11.

100

 

 

 

 

 

 

 

 

 

 

 

 

2.5 . 2
2 EA D B :
1 P
= :15
1.5—3 0 I E
1.: U . E"

s I i
0.51 I . :‘0-5
i D I I 0 E
415-: B La;

4.; 5
a 0 E4
4.51 :
'2'; $4.5
8 .205 ‘ T—V—UTrWI t1 1 T I Y I I TIT It I V I I II V ‘T VTV'TIT'YIT'YII'U'VIVVIYIIYI‘r“'t‘v'frb .2
°- 2 q 2
: c o D E
1.5-3 '—
5 Cl :15
1-3 0 91
11.5-f I . . £115
3 . O :
-o.5-j D I 0 o {—41.5
3 C].- DE
-‘1 fl :7"

.1 I
4.51 E1451
4 1-

.2:1 h
I v IT I I I I I I I I I l I l V 1“ | I ' rtvrItvtt'ItrrIVYITIVT1YItYfrlrtYVI—IIT! 2
'1 .5 '1 '05 “-5 I 1-5 -2 4 .5 -I ~0.5 0 0.5 1 I .5 2

Figure 4.6. PCA ordination of bacteria peak area data for leaf (A and B) and water (C

PC1

and D) samples taken from Block I (A and C) and Block 11 (B and D) microcosms on day

20. The amount of variation explained by each PC is indicted on the axes for each panel.

Treatments are represented by the following symbols: open, 0 scirtids; filled, 10 scirtids;

squares, 0 mosquitoes; circles, 40 mosquitoes.

101

 

 

 

 

 

 

 

2 I
5A B 52
1.5-3 0. 54.5
1.: . . 0 E4
0.5—5 % I g D g—o.5
0-3 0 §.o
: .0 g :
41.51 0 8 :--0.5
-l—':‘ :..1
~1-5:§ E—-1.5
41 O 5‘2
'25“: . E--2.5
8 -3 ' [r1 Tl fi—VIUIVTT' 1‘" I III‘r'TVIIVIIIIIIIY'II‘TY'IIYU‘VY‘VAIIITI '7‘ 3 g
“- 2. _ 2 N
:c D :
15~j :1 -15
1'3 0 :4
E I E
0.51 D G 70.5
E I E
0': I (9 C] I 1-0
3 o
41.5: - ;--115
-1- o O o L.
. U o .
4.5-5 . 54.5
: . 9
-2 IIIIIYTT‘IIYIIII'YYTIIIYV'I'I'Ijj' IIYI'IT'UITYY‘I'ITW"! IV'IIT'TTIIIIYT'TI S

 

 

 

 

up.
-2 -l.5 -1 41.5 0 0.5 I 1.5 -2 4.5 -1 -0.5 0 0.5 I 1.5 2 2.5
PC1

Figure 4.7. PCA ordination of fungal peak area data for leaf (A and B) and water (C and
D) samples taken from Block I (A and C) and Block 11 (B and D) microcosms on day 20.
The amount of variation explained by each PC is indicted on the axes for each panel.
Treatments are represented by the following symbols: open, 0 scirtids; filled, 10 scirtids;
squares, 0 mosquitoes; circles, 40 mosquitoes.
Discussion

Although scirtids had major effects on microbial community parameters, this did
not translate into positive affects on mosquito populations expected if facilitation is
occurring. While it is possible that the lack of scirtid-mediated facilitation in this study
may be an artifact of our experiment, it is more likely that any positive scirtid impact may

have been negated by competition among the macroinvertebrates for the same resources.

102

Other studies have indicated that scirtid beetles may exert negative effects on mosquito
populations, such as decreased body mass at adulthood and reduced survival, when
resource levels are high (Paradise 1999, 2000). This switch from commensalism to
amensalism under different nutrient regimes, also described by Heard (1994) in another
processing chain commensalism, may manifest as a continuum of affects on mosquito
performance, ranging from highly facilitative to negative. Thus, we postulate that our
current findings represent a condition under which resources were sufficiently available
to both macroinvertebrates, resulting in a neutral affect of scirtids on mosquito
populations.

This study evaluated the impact of scirtid beetle presence on microbial
community dynamics in tree holes. We demonstrated that even at low densities, scirtid
beetles increased the rate of decay of leaf detritus and altered the community structure of
microorganisms in water and on leaves. Although leaf fungal biomass was slightly higher
but not significantly so in the presence of scirtids, lower remaining leaf mass in these
treatments means that on a per mass basis, a greater percentage of fungal biomass was
contained in leaf material in the presence of scirtids. This conclusion is supported by
corresponding increases in the activity of the leaf polymer degradation enzymes, xylanase
and cellobiosidase, in association with scirtid beetles. Taken together, these results
indicate accelerated leaf decay associated with beetle feeding activity, supporting the
hypothesis that increased processing of leaf detritus drives facilitation. Although scirtid-
induced reductions in leaf mass have been documented previously (Paradise 2000), the
current study is the first to indicate a microbial mechanism for leaf processing other than

physical conversion of coarse particulate organic matter (CPOM) to fine particulate

103

organic matter (FPOM). In addition, although the latter contributes to the pool of
resources utilized by 0c. triseriatus, transitions between filtering (consuming FPOM and
microorganisms) and browsing (consuming microbial biofilms) feeding modes are known
to occur (Menitt 1992). If browsing behavior is indeed more commonly exhibited by
late-instar 0c. triseriatus, then we would expect that late-instar mosquito larvae would
benefit fi'om the increase in leaf-associated fungal biomass conferred by scirtid beetles.

Presence of scirtid beetle larvae was insufficient to reduce bacterial production
associated with leaf surfaces, although the effect of scirtids was marginally significant. In
contrast, the mosquito effect on leaf productivity is pronounced. A significant interaction
between mosquito and scirtid main effect indicates that mosquitoes drive down
productivity even in the absence of scirtids. That mosquitoes drive down bacterial
productivity on leaves is well-described by Kaufman et al. (1999, 2001, 2002, 2006);
however, this is the first study to indicate that another tree hole invertebrate may also
contribute to reduced productivity. Although a reduction in bacterial productivity in the
presence of mosquitoes also occurred in the water column, a corresponding change in
bacterial abundance did not occur in response to the presence of either mosquitoes or
scirtids. This is unsurprising in light of previous findings from tree holes. Kaufman et al.
(2001) suggested that such decoupling of bacterial abundance and growth rate may result
from shifts in the structure of bacterial communities in response to macroinvertebrate
presence.

While scirtids did not increase the abundance or productivity of water column
bacteria, they did have an impact on water column bacterial community structure. Using

T-RFLPs, we determined that when scirtid beetles are present shifts in community

104

structure occur involving a few key bacterial groups. Bacterial community shifts in
response to mosquito larvae have been described for pitcher plant and tree hole
communities (Cochran-Stafira and von Ende 1998, Kaufman et al. 1999, Kneitel 2002);
however this is the first study to evaluate this response in the presence of scirtid beetles.
In contrast to bacterial communities, leaf surface fungal communities changed in
response to time and mosquito larvae, but not the presence of scirtid beetle larvae.
Although communities were more productive and had greater biomass (ergosterol),

neither scirtids nor mosquitoes changed the nature of this community.

That bacterial community changes evident on the leaf surface were influenced by
scirtid as well as mosquito presence is unsurprising because scirtid feeding would
necessarily result in the consumption of surface-associated bacterial biofilms along with
leaf particulates. However, the effect of scirtids on the water column bacteria suggests
their movements (feeding and locomotion) along leaves may dislodge surface bacteria
and FPOM into the water, thus altering the bacterial community in both locations. Indeed,
the same effect is likely caused our water-column fungal communities to shift in response
to scirtids. Overall, mosquito presence had the most influence on changes in microbial
community composition. Surface and water column —associated bacterial and fungal
communities all changed significantly in response to mosquito presence, in
correspondence with our previous observations (Kaufman et al. 1999, 2008, Xu et al.
2008) Although we did not assess water column fungal biomass in the present study, we
suspect that mosquitoes change fungal composition in the water column by reducing the

taxa susceptible to larval digestion (Graca 2001, Rossi 1985).

105

The community results obtained using T-RFLPs are limited to the level of
classification conferred by fragment sizes. Although differences might have not been
apparent based at the “species” levels, defined herein as fragments differing in size by at
least one base pair, changes in community structure might be more apparent at higher
orders of classification. Surprisingly, though scirtids may open additional niches for
fungal colonization via shredding leaf material, scirtids did not appear to directly affect
the structure of fungal communities through their feeding activity. The absence of scirtid-
associated changes in the structure of fimgal communities may indicate these
macroinvertebrates are not affecting the structure at all, or at least not at the “species”
level of resolution provided by T-RFLPs. Alternatively, changes may not be visible as a
result of the technique used to harvest surface associated microorganisms from the leaf
surfaces. Loosely attached fungi are represented in the sonicated leaf samples while fungi
whose hyphae are enmeshed in the matrix of the decomposing leaf are not. Rather than
accessing only loosely attached fungi, beetles consume whole leaf particles with the
associated fungi; therefore, changes in the composition of fungal communities may be
relatively uniform, in contrast to the compositional changes in response to mosquitoes
observed in this study and others (Kaufman 2008). Thus, scirtids may very well be
altering leaf-associated fungal communities and increasing fungal-derived nutrients in the
system via feeding on embedded fungal hyphae, however, that question should be
addressed in future experiments.

We found that the two enzymes used for TRFLP digestions in some cases
produced slightly different results that, although not uncommon, underscore the necessity

of using multiple enzymes to refine the “species” composition obtained for each

106

treatment. To our knowledge, this is the first study to utilize T-RFLP analysis for
assessing microbial community changes in a replicated experimental study of mosquito
habitats. While previous studies have had success utilizing cloning and other methods of
microbial community structure analysis, these techniques are often laborious and limited
by the number of replicate samples that may be reasonably, and economically, processed.
In addition, differences among microbial community results in blocks I and 11 suggest
variation in the initial inocula used in microcosms. Although composite inocula were
used in both cases, the inocula were obtained at separate times, immediately preceding
the onset of each experiment. The compositional differences in these communities
underscore the heterogeneity present in tree hole communities and deserve consideration
in future studies of scirtid and mosquito populations. Variation in the condition of
available resources is postulated to underlie the stochastic distribution of scirtid beetles
occupying nearby tree hole such that tree holes with high quality litter resources are
likely to harbor scirtid populations (Paradise 1998; Paradise & Kuhn

1999; Paradise 1999).

As shredders, scirtid feeding in tree holes accelerates detrital processing by
reducing leaf material into finer fragments. From a mosquito’s perspective, scirtid
feeding alters detrital surfaces and water column content, thus directly impacting the
primary larval feeding zones. Leaf surfaces in particular are a tough, non-nutritive vehicle
upon which resides a film of rich nutrients incorporated as microbial biomass. Although
bacteria in this biofilrn are severely grazed down by mosquito larvae, fungal members of
this biofilrn are less susceptible because they are anchored into the matrix of decaying

tissue itself. Scirtid feeding may make this resource more available to larvae by exposing

107

hyphae and/or releasing small, ingestible fragments containing fimgal hyphae into the
filter feeding zone. The fact that scirtid feeding altered water column fungi supports the
latter mechanism, but measurements of fimgal biomass in the water column are necessary
to firrther this hypothesis. Because resource availability is the most important
contributing factor to mosquito growth, scirtid-associated changes in the underlying
structure of microbial communities and the spatial rearrangement of these communities
may provide mechanisms by which scirtid beetle presence in tree holes alters mosquito

success under nutrient-limited conditions.

108

CHAPTER 5
COMPETITIVE INTERACTIONS BETWEEN OCHLEROTA TUS TRISERIA TUS
(SAY) AND AEDES ALBOPICTUS (SKUSE) (DIPTERA: CULICIDAE) ARE
INFLUENCED BY HABITAT-ASSOCIATED MICROBIAL COMMUNITY

DYNAMICS

109

Introduction

Container-dwelling forms represent approximately 40% of mosquito species,
many of which are important arbovirus vectors, including West Nile Virus, yellow fever
and LaCrosse encephalitis (Laird 1988). The vectorial capacity of adult mosquitoes is
directly related to the number of adults produced from the natal container habitat. Under
conditions of high larval density, competition for food resources is intense and inherently
changes the availability of such resources. Limited food resources at the larval stage
results in the production of smaller adults with lower metabolic reserves, and leads to
higher nutrient demands at this stage and thus increased incidence of host contacts. As a
result, these mosquitoes may be more competent vectors of arboviruses (Alto et al. 2005).
Several studies have shown that the Asian tiger mosquito, Aedes albopictus, is superior to
the Eastern tree hole mosquito, Ochlerotatus triseriatus, under conditions of larval
competition. Both species are major vectors of human pathogens (West Nile Virus and
LaCrosse encephalitis, respectively).

Competition, both inter— and intraspecific, is one important factor mediating the
production of adult mosquitoes within and among species sharing water-filled container
habitats. Interspecific competition is common and occurs depending on the range overlap
of container-breeding mosquitoes (Kitching 2001, J uliano and Lounibos 2005). In
southern latitudes, Aedes albopictus and A. aegypti fi'equently share larval habitats (Rai
1991). Ochlerotatus (= Aedes) triseriatus may overlap with either of these species in the
southernmost portion of their range, which extends to Florida in the Eastern United
States. In its northern range, 0c. triseriatus has historically been free from intraspecific

competition; however in recent years Ochlerotatusjaponicus has begun to invade

110

container habitats in this region (Sardelis et a1. 2001, Scott et al. 2001, Roppo et al.
2004)

Compared to larger water body mosquito habitats, density dependent mortality
unrelated to predation appears to be a primary population regulator of container-dwelling
larval mosquitoes as species compete directly for the unpredictable food resources that
characterize container habitats (J uliano 2007). Container-dwelling mosquitoes belong to
the grazing/filtering functional feeding group; larvae depend on the microbial biofilm
associated with leaf and container wall surfaces, and on fine particulate organic matter
(F POM) and planktonic microbes present in the water column (Walker and Merritt 1991).
Interspecific competition functions among cohorts of larvae through shared resource
pools and shared food acquisition behaviors (Kitching 2001). Within a single species, we
expect that food acquisition behaviors are similar in terms of the proportion of time spent
browsing or grazing on surfaces verses filtering in the water column or at the water
surface. Among competing species, these proportions may differ, as is the case between
A. aeqypti and A. albopictus. For example, when leaves are the food substrate A.
albopictus outcompetes A. aegypti (Barrera 1996, J uliano 1998, Daugherty et al. 2000,
Braks et al. 2004, Yee et a1. 2004), however the advantage shifis to A. aegypti when
animal carcasses are utilized as a food resource (Daugherty et al. 2000). Yee et al. (2004)
suggest that the dominance of A. albopictus over A. aegypti in all but the latter case can
be attributed to differences in the relative time spent browsing on leaf substrates. In their
study, A. aegypti spent more time filter-feeding in the water column, where microbes are

less abundant.

111

Previous studies have examined interspecific competition among mosquito larvae
by measuring adult productivity at different resource levels when larvae are alone or
under interspecific competition. In addition to the above studies of A albopictus and A.
aegypti, differences in success at low resource levels are also seen in competition studies
involving 0c. triseriatus and A. albopictus (Livdahl and Willey 1991, Novak et al. 1993,
Teng and Apperson 2000). Although the proximate cause of this outcome is unknown, it
is apparent that A. albopictus has the potential to replace 0c. triseriatus where the species
co-occur. These studies suggest a clear competitive advantage for one species at a defined
low resource level or as detritus sources vary (Yee et al. 2007). However, given our
current understanding of how mosquito larvae utilize resource pools, the mechanism
driving such outcomes remains ambiguous. The lack of a clearly-defined mechanism
remains an important gap in our understanding of asymmetric success under competition.

Complicating the ambiguous mechanism is the ill-defined nature of the food
resources in larval mosquito habitats such as tree holes. Allochthonous inputs, consisting
of plant and animal detritus, and stemflow are the primary nutrient-supplying inputs into
the system (Carpenter 1982, Daugherty et al. 2000, Kitching 2001). Plant material in the
form of senescent leaves appears to supply the majority of the organic carbon to these
systems and is thus the resource input typically used by researchers for laboratory-based
microcosms simulating natural habitats. Leaf matter, as stated above, cannot be ingested
directly by larvae but is accessed instead via intermediate microbial processing in which
nutrients from leaf material and stemflow inputs become available to larvae as they are
incorporated into microbial biomass. Softening of the leaf tissue through microbial

processing may also contribute to an accessible pool of FPOM, but the extent to which

112

this may occur is unknown. Therefore, it is apparent that any study of interspecific larval
competition must take into account the microbial community that comprises the actual
resource for which species compete.

Studies suggest that larval feeding alters the microbial communities present in
tree holes, resulting in bacterial forms that are indigestible, and thus unavailable, to the
mosquito larvae (Kaufman et al. 1999, 2008, Xu et al. 2008). Furthermore, several groups
of microorganisms, including protozoa, disappear or are severely depleted from the
system under larval feeding pressure compared with larva-free conditions (Kaufinan et al.
2002). It is therefore probable that the mechanism underlying competition among larvae
is the depletion of available consumable microbial forms, rather than direct utilization of
detrital resource pools. We tested the hypothesis that competing mosquito species
differentially utilize the available microbial resources to produce the differential grth
parameters observed under competition. We postulate that one of the following
mechanisms are responsible for the result of interspecific competition among larvae: 1)
species exhibit differences in their ability to utilize available microbial forms, 2) within-
species feeding patterns result in differential harvesting of microbes from resource
reservoirs i.e. water column verses leaf surfaces, and 3) one species may drive down the
total available microbial resources to a level below which the competing species cannot
maintain comparable productivity. The purpose of this study was to assess the structure
of the microbial community associated with tree hole habitats to determine the viability
thereof as a resource mediating interspecific competition among larvae. Specifically, our

objective was to evaluate known competitive interactions between A. albopictus and the

113

Eastern tree hole mosquito, 0c. triseriatus, to determine whether the documented success
of A. albopictus derives from one of the three mechanisms described above.
Materials and Methods

Microcosms. We constructed microcosms containing 0c. triseriatus and A.
albopictus under interspecific and intraspecific competition at two mosquito densities.
Senescent red oak leaves (Quercus rubra L.) collected at Michigan State University’s
Kellogg Forest (Augusta, M1) were dried at 45°C for 48 h. Treatments were applied to
individual microcosms stocked with 1.0 g oak leaf packs, 500 ml deionized water, and a
microbial inoculum consisting of 3 ml homogenized natural tree hole water and
particulates.

Prior to the addition of newly-hatched first instar mosquito larvae, all microcosms
were conditioned for 3 days to allow time for microbial colonization of leaf surfaces and
water column. The larvae used in the following experiments were hatched from eggs
collected from our colonies in the Insect Microbiology laboratory at Michigan State
University.

To evaluate the dynamics of microbial resources underlying interspecific
competition between 0c. triseriatus (T) and A. albopictus (A), microcosms were treated
with two levels of two factors, competitor (intraspecific, or TT and AA, vs. interspecific,
or AT) and density (high vs. low), for a total of a total of six treatment combinations.
Intraspecific competition microcosms contained an equal ratio of individuals from each
species i.e. 15:15 or 30:30.To evaluate the effect of density on mosquito performance,

initial densities of 60 (high) or 30 (low) first instar larvae were added to microcosms.

114

Treatment combinations were replicated a total of twelve times to provide sufficient
microcosms for destructive sampling.

Samples. Microcosms were sampled four times during the 60 day course of the
experiment- at day 0 (prior to larval addition), day 15, day 30 and day 60. On each date,
four replicate microcosms fi'om each treatment were destructively sampled.

During each sampling period, leaf discs and of water samples were removed from
microcosms for measuring productivity, abundance, and structure of microbial
communities as described below. Although the data are not presented here, additional
water samples were taken from microcosms on all sampling dates for nutrient (nitrogen
and phosphorus) analysis. Leaf samples were aseptically procured with a cork borer (11
mm diameter), placed in filter-sterilized microcosm water (productivity — see below) or
phosphate-buffered saline (abundance and community analysis — see below), and
sonicated for 12 min to dislodge loosely-associated microorganisms.

Mosquito Measurements. Microcosms were checked daily for the presence of
adult mosquitoes. Adults were collected and stored at 4°C until the microcosm was
destructively samples. Larvae were gathered by decanting the remaining microcosm
water through a fine mesh sieve and stored at 4°C. Where possible, we sexed and
identified adults, larvae, and pupae to species before freeze-drying the mosquitoes to
obtain dry mass measurements.

Bacterial Abundance. Leaf- and water column-associated bacterial abundance
was assessed by removing a single leaf disc and 15 ml of water from microcosms per
sampling period. Samples were preserved for later analysis with formalin at a final

concentration of 3.4%. Bacterial abundance on the leaf surface and in the water column

115

subsamples was quantified via direct microscopic counts (DMCs) of bacteria and
protozoa using the DAPI (49,6—diamidino-2-phenylindole) fluorescent staining procedure
(Porter and Feig 1980, Walker et al. 1988). Samples were filtered onto two black
Nucleopore filters (0.2-mm pore size; Costar, Cambridge, Mass), stained at 2 mg/ml for 5
min, and counted at 1000X.

Bacterial Productivity. A 3H-leucine incorporation assay was used to directly
measure accumulation of microbial biomass (Kirchman 2001). Quantifications of the
amino acid incorporated into protein are achieved in this technique in a bacteria-specific
manner using short incubation periods of samples with nanomolar concentrations of
leucine (Riemann and Azam 1992). Labeled leucine (L-leucine (4,5-3H ), 50 Ci/mmol—
NEN, Life Science, Boston, MA) was added to water and sonicated leaf samples at a
concentration of 25 nM. To achieve previously reported saturation of uptake kinetics
(Kirchman 2001, Kaufman et al. 2001), unlabeled leucine was added with labeled leucine
to bring concentrations in water and leaf samples to 100 and 400 nM, respectively.
Leucine incubations were done in the dark at 20°C for 20 min in 2 m1 microcentrifuge
tubes (Smith and Azam 1992, Kirchman 2001). Trichloroacetate [TCA, final
concentration 10% (vol:vol)] was added to terminate reactions and precipitate protein.
Two rinses of the TCA-protein precipitates were conducted with 10% TCA, followed by
a single rinse with 5°C, 80% (vol:vol) ethanol. A Beckman LS6500 scintillation counter
(Beckman Coulter, Inc., Fullerton, CA) was used to quantify the amount of radioactivity
present in the samples.

T-RFLP Analysis. The structure of leaf- and water column-associated bacterial

and fungal communities in microcosms was assessed using tagged restriction fragment

116

length polymorphism (T—RFLP) analysis (Marsh 1999). We extracted DNA from the
water column and sonicated leaf samples described above using the MoBio UltraClean
soil extraction kit (Carlsbad, CA). Universal primers targeting segments of the small
subunit (SSU) ribosomal DNA and specific to bacteria and fungi (16S rDNA and18S
rDNA, respectively) were used for amplifications. Bacterial DNA was amplified using
universal eubacterial primers 63F (5’-CAGGCCTAACACATGCAAGTC-3’) and 1387R
(5’-GGGCGGWGTGTACAAGGC-3’) (Marchesi et al. 1998). Fungal DNA was
amplified using primers nu—SSU-O817-59F (5 ’-TTAGCATGGATAATRRAATAGGA-
3’) and nu-SSU-1536-39R (5’- TTGCAATGCYCTATCCCCA—3’) (Bomeman and
Hartin 2000). The forward primers were fluorescently~labeled at the 5’ end with FAM
(carboxyfluorescein) for detection by capillary electrophoresis. Reaction mixtures (100 111
final volume) included ca. 10 ng DNA template, 50 ul Failsafe PreMix buffer ETM
(Epicentre Biotechnologies, Madison, WI), 1 ul FailsafeTM enzyme, and 4 ul forward and
reverse primers. The cycles used for 16S rDNA were as follows: 1 cycle at 94°C for 2
min; 30 cycles at 94°C for 45 s, 68°C for 30 s, and 72°C for 1 min 30 s; andl cycle at
72°C for 7 min. Similarly, 18S rDNA was cycled through the following: 1 cycle at 94°C
for 2 min; 30 cycles at 94°C for 45 s, 58°C for 30 s, and 72°C for 1 min; andl cycle at
72°C for 7 min.Gel electrophoresis was performed to verify amplification of template
DNA. Three PCR reactions per sample were pooled and purified (Qiaquick PCR
purification Kit, Valencia, CA). PCR products were digested independently with two
restriction enzymes, MspI and leal (New England Biolabs, Cambridge, MA) at 37°C for
3 h. Following electrophoresis of digested samples on a capillary electrophoresis genetic

analyzer (PRISM 3100 Genetic Analyzer, Applied Biosystems, Foster City, CA), the size

117

and frequency of each terminal restriction fragment was determined using Genescan
Analysis software (Applied Biosystems).

Statistical Analysis. Multivariate analysis of variance (MANOVA) was used to
analyze related variables: water chemistry values, mosquito productivity values, bacterial
productivity, and bacterial production (SAS system fir Windows 9.1; SAS Institute, Cary,
NC). Mosquito productivity values included the mass, survival, and total emerged adults
(males and females) obtained per microcosm. Peaks in T-RF electropherograms were
identified and binned using the T-RFLPS Stats tools (Abdo et al. 2006) and R (R
Development Core Team 2004). The relative percent abundance of peak areas obtained
from T-RFLP profiles obtained fi'om bacterial and fungal communities were transformed
(logratio) then subjected to principal component analysis (J MP® Statistical Discovery
Software, V5.1 (http://www. jmpin.com, SAS Institute, Inc., Cary NC, USA).

Results

Mosquito production. Mosquito species and density significantly affected 0c.
triseriatus populations, but not A. albopictus (Table 5.1), suggesting interspecific
competition between the two species. The success of 0c. triseriatus (adult body mass,
adult emergence, and overall survival) was lower under interspecific competition at high
total mosquito densities, as indicated by a significant interaction between mosquito
density and species (Table 5.1 and Figures 5.1-5.3). More A. albopictus adults were
produced than 0c. triseriatus under every condition, although individual species
performed better at low intraspecific densities (Figures 5.1-5.3). In addition, A. albopictus
emerged sooner than did 0c. triseriatus for both density and competition levels. In low

density treatments, pupation ranged from day 13 for intraspecific A. albopictus to day 34

118

for 0c. triseriatus (Figure 5.1). Microcosms were preserved until day 60, although
additional pupation did not occur beyond day 34. For both species male mosquitoes
developed earlier than females under inter- and intraspecific competition. In addition, the
mean adult emergence was significantly earlier for A. albopictus that 0c. triseriatus.
More A. albopictus adults were produced than 0c. triseriatus when the two species
competed and under low density intraspecific competition (Figure 5.2, C and D). At high
densities, fewer mosquitoes emerged under interspecific conditions compared with
intraspecific conditions, regardless of species. The opposite trend occurred for A.
albopictus, which exhibited increased adult emergence when competing with 0c.
triseriatus at low density.

Under intraspecific competition, the mean body mass was similar among males of
both species, however under interspecific competition, the mean body mass of A.
albopictus males increased relative to 0c. triseriatus (Figure 5.3, A and B). In contrast,
the body mass of male 0c. triseriatus under interspecific conditions was depressed or
unchanged compared to intraspecific levels. Female 0c. triseriatus body mass remained
the same under low density interspecific conditions, but could not be measured at high
densities, as no adult females emerged from these microcosms Figure 5.3, C and D).
Reductions in adult emergence and body size when individual species were at high
densities suggest that intraspecific competition was more intense than at low densities.

Bacterial abundance and productivity. Bacterial abundance in the water
column and on leaf surfaces did not differ significantly between mosquito densities or
under inter- or intraspecific competition treatments (Figure 5.4; Table 5.2). Over time, the

abundance of bacteria remained the same, although the number of cells present in the

119

water column were consistently lower than on leaf material (Scheffe' a posteriori test, P
= 0.05).

Productivity in the water column and on leaf surfaces decreased significantly
from day 15 to day 30 (Figure 5.5; Table 5.3). In both homogenous and heterogeneous
mosquito populations high larval densities were significantly associated with reductions
in water column bacterial productivity compared with low density treatments, although
this effect was mitigated by time. Despite the significant interaction of larval presence
and sampling date, the species combination treatments did not exert a significant effect
on the production of bacteria on any of the sampling dates examined. On days 30 and 45,
there was no difference in productivity between the two larval densities, nor did
productivity differ in the treatments between the two sampling dates (Scheffe' a

posteriori test, P = 0.05).

120

Table 5.1. Multivariate analysis of variance (MAN OVA) results for emergence and mass
of A. albopictus and 0c. triseriatus in microcosms. Total adults and females were

analyzed as separate dependent variables.

 

 

Source *F statistic df P value
A. albopictus

Mosquito 1.4 5,6 0.34

Density 2.65 5,6 0.13

Mosquito*Density l .89 5,5 0.25
0c. triseriatus

Mosquito 0.13 5, 9 0.001

Density 0.33 5,9 0.05

Mosquito*Density 0.23 5, 8 0.02

 

*Wilkes’ Lambda.

121

A. albopictus O. triseriatus

 

 

 

 

 

 

 

35 35
A B
30 i 30
25 § g 25
§~ 20 i i ,20
3.; 15 < 15
‘5
.5, 1o 10
LI.
5 5
o o
35
c o 35
30 f 30

 

 

 

 

 

5 25 g § 25

e

3 20 20

. i

2 15 .15
1o 10
5 5
o o

lntraspecific lnterspecific lntraspecific Interspecific

Figure 5.1. Development time for two densities of A. albopictus and 0c. triseriatus in
microcosms. Low mosquito densities are represented by filled characters and high
densities by open characters. Panels A and B are females, C and D are males. Values are

means :1: SE, n = 3-4 for each point.

122

A. albopictus O. triseriatus

 

 

 

 

 

 

 

 

 

 

 

 

 

 

12 5
A B
l
" 10 f
b”, 4
2
a 8 i
g -3
3 6
E
2
i9 4 i
.9
.2
2. * 1
0 :1 [l a o
30 14
C D
25 12
a l
10
4:4 20
3 a
g 15»
.03
3 6
«I 10
E 4
,2
5 . i? Q i , 2
o . 3 o
lntraspecific lnterspecific lntraspecific Interspecific

Figure 5.2. Number of adult A. albopictus and 0c. triseriatus emerged from microcosms.
Low mosquito densities are represented by filled characters and high densities by open
characters. Panels A and B are females, C and D are total number of adults. Values are

means :t SE, n = 3-4 for each point.

123

A. albopictus O. triseriatus

 

 

 

 

 

 

 

 

 

 

 

 

 

3.5 1.25
A .. B
3
' 1
2.5
i
’6: 2 0.75
E,
«11.5
g 0.5
1
05 E] 025
ll
0 D J I] r o
3
C D
5 -- 2.5
g’ 4 2
_a_>
E 3 - 1.5
a: l
2 1
1 Q [E 0.5
. ° '1
0, {3 o
lntraspecific Interspecific . lntraspecific Interspecific

Figure 5.3. Mass of adult A. albopictus and 0c. triseriatus from microcosms. Low
mosquito densities are represented by filled characters and high densities by open
characters. Panels A and B are females, C and D are males. Values are means :1: SE, n =

3-4 for each point.

124

Table 5.2. Multivariate analysis of variance (MANOVA) results for bacterial abundance

in microcosms on days 15 and 30.

 

 

F statistic df P value
Bacterial abundance
Mosquito 0.84 8, 30 0.57
Density 1.67 4, 15 0.21
Mosquito*Density 1.05 8, 30 0.42

 

*Wilkes’ Lambda.
Table 5.3. Multivariate analysis of variance (MANOVA) results for bacterial

productivity in microcosms on days 15, 30 and 60.

 

Source F statistic df P value
Time 1.6 8, 98 0.1358
Mosquito 0.6 4, 98 0.6627
Density 9.24 2, 49 0.0004
Mosquito*Time 6.86 4, 98 <.0001
Density*Time 1.34 8, 98 0.2322
Mosquito*Density 1.75 4, 98 0.1451
Mosquito*Density*Time 32.2 4, 98 <.0001

 

*Wilkes’ Lambda.

125

Cells x 107/ ml or mm2 (mean t SEM)

Water Column

Leaf Surface

 

i

4
1
1
i
l
1
1

A

C

 

 

111i
E11.- .. fl ...... I1 ++++++++ [1

Competition Treatment

Day 15

Day 30

Figure 5.4. Bacterial abundance (direct microscopic counts) in microcosms containing

two densities of A. albopictus (A) and 0c. triseriatus (T), alone or under interspecific

competition. Low mosquito densities are represented by filled characters and high

densities by open characters. Panels A and C are water column samples, B and D are leaf

surface samples. Samples were taken on day 15 (A and B) and Day 30 (C and D). Values

are means iSE, n = 3—4 for each point.

126

6.0

4.0

2.0

2.0

1.6

3H-Ieucine Incorporation nmol x 104/ ml/ h
0

1.2

0.8

0.4

Day

15

Day 30

 

 

 

 

 

 

 

 

 

 

 

M

 

 

 

 

 

AT

 

TT

 

 

 

ME!

Competition Treatment

Water column

Leaf surface

Figure 5.5. Bacterial productivity (leucine incorporation rates) in microcosms containing

two densities of A. albopictus (A) and 0c. triseriatus (T), alone or under interspecific

competition. Low mosquito densities are represented by filled characters and high

densities by open characters. Panels A and B reflect samples taken from the water

column, C and D are from leaf surfaces. Samples, taken on day 15 (A and C) and day 30

(B and D), are represented as means i SE, n = 3-4 for each point.

127

T-RFLPs. Water-associated bacterial communities exhibited significant
differences along principle component 1 on days 15 and 30 in response to mosquito
density, but not in response to interspecific or intraspecific mosquito species combination
(Figure 5.4; Table 5.2). Such differences were also apparent in leaf bacterial communities
on day 15, but not on day 30. Neither mosquito nor density exerted a significant effect on
PC2 (data not shown). Cumulatively, PC 1-3 explained 38-79% and 67-78% of the
overall variance in leaf and water column associated bacterial communities, respectively
(Figure 5.6). Restriction fragment 109 accounted for most of the variability in the water
column bacterial community along PC1 and fragment 398 along PC2; however, the
abundance of fragment 109 decreased from day 15 to day 30 (Figure 5.6; Table 5).

In contrast to bacterial communities, mosquito treatment significantly interacted
with density to influence water column fungal communities on day 15 along PC1 (Figure
5.7; Table 5.3). This interaction effect was also slightly evident on 30 at P = 0.1. A
significant affect of mosquito species combination on the structure of leaf-associated
fungal commmrities was reflected by PC2 on days 15 and 30 (Figure 5.7; Table 5.4).
Water column fungi were also influenced significantly by mosquito treatment on day 30,
however this effect was not evident on day 15. The total variation leaf and water column
fungal communities explained by PC1-3 ranged from 47-54% and 54-64%, respectively
(Figure 5.7). Although not shown here, all differences in the structure of microcosm
microbial communities associated with mosquito species competition and mosquito

density had disappeared by day 60.

128

Table 5.4. Analysis of variance (ANOVA) results for PC1 values obtained from PCA

analysis of bacterial (l6S rDNA) T-RFLP peak area data.

 

Water Leaf
Source F statistic df P value F statistic df P value
day 15
Mosquito 0.50 2 0.62 0.58 2 0.57
Density 17.50 1 0.001 5.65 1 0.03
Mosquito*Density 1.29 2 0.31 1.36 2 0.29
Error 14 13
day 30
Mosquito 1.03 2 0.41 1.07 2 0.36
Density 8.66 1 0.02 1.88 1 0.19
Mosquito*Density 0.21 2 0.82 1.98 2 0.17
Error 6 l7

 

129

 

 

 

 

 

 

 

 

 

 

1; A B ; T. t
. o O I
I I - Q 0 I
0.5 1 ‘ . 0 * _— 05
"‘ . a. - I
d ‘ r P
i h D ..
0" I f 0
: 0 ~
~ I O A I
-o.5 1 ~ -o.5
1 D O ‘
-1 - . — B :‘ ‘1
3 C1 2
8 -15 I'YIIYYYVIVUU'IIU'YI‘IIY[VYUYI'IIVII{I‘ll WI‘IIIVYIVIYFI'IIUII'U'U'I'TY'ITIIIlIYIY -'.5 a
n. 2 . c D r 2 N
5 O I :_
: .. o :
i O "" C
E “ :3 :
“z ‘- I r
5 g 3 O — a o :
01 A _ 5] :- °
.3 A A I A r
E e - 0 E
-1 1 I — :4
j 0 E
E “ I E
.2 ‘TTTr'T'Y1r'IT'IIII'UYIIIIYIIIUIIIT‘T'II" III'lIfi‘I'I'I'IIIIIUIIUIIVYIIVYWIIIrWYI .2
.2 -1 o 1 2 .2 -1 o 1 2

Figure 5.6. Principle component analysis ordination of bacterial T-RFLP peak area data
for microcosms containing two densities of A. albopictus (squares) and 0c. triseriatus
(circles), alone or under interspecific competition (triangles). High mosquito densities are
represented by filled characters and low densities by open characters. Panels A and B are
water column samples, C and D are leaf surface samples. Samples were taken on day 15

(A and C) and Day 30 (B and D). n = 3-4 for each point.

130

Table 5.5. Analysis of variance (ANOVA) results for PC1 values obtained from PCA

analysis of fungal (18S rDNA) T-RFLP peak area data.

 

Water Leaf
Source F statistic df P value F statistic df P value
day 15
Mosquito 0.50 2 0.62 1.60 2 0.23
Density 0.30 1 0.59 4.21 1 0.06
Mosquito*Density 4.29 2 0.03 0.38 2 0.69
Error 14 17
day 30
Mosquito 3.51 2 0.05 3.05 2 0.08
Density 0.26 1 0.61 2.08 l 0.17
Mosquito*Density 2.92 2 0.08 0.38 2 0.69
Error 15 15

 

131

Table 5.6. Analysis of variance (ANOVA) results for PC2 values obtained from PCA

analysis of fungal (18S rDNA) T-RF LP peak area data.

 

Water Leaf
Source F statistic df P value F statistic df P value
day 15
Mosquito 0.61 2 0.56 6.08 2 0.01
Density 0.22 1 0.23 0.54 l 0.47
Mosquito*Density 0.63 2 0.63 0.04 2 0.96
Error 14 17
day 30
Mosquito 7.32 2 0.01 5.39 2 0.01
Density 0.60 1 0.45 0.42 1 0.53
Mosquito*Density 3.14 2 0.07 0.04 2 0.96
Error 15 15

 

132

Table 5.7. Bacterial (168) T-RFLP fragments with high factor loadings.

 

 

Day 15 Day 30
Substrate PC1 PC2 PC1 PC2
Water 109, 451 365, 398 109 252, 398, 456
Leaf 456, 460 258, 456 118, 150, 394 150, 394, 597

 

Table 5.8. Fungal (18S) T-RF LP fiagments with high factor loadings.

 

 

 

Day 15 Day 30
Substrate PC 1 PC2 PC1 PC2
Water 582, 767 582, 559, 767 211, 547, 707 559, 760
Leaf 212, 559, 767 527, 559, 592 53, 559, 584 53, 584, 716, 767

 

133

 

 

9N
U!

rrrrlrrrrlrrrrlrvrr
F, ct
U!

o
>
9i

q
‘
rtrrrrrrr'vrrr
I
I.

 

 

P02

N
N
ZOd

 

 

O
U

a.

in
ii ILJJ

r] t 1

I.

in

If
I.

IIIIJILIL
D
l
m ‘I
‘O

O
.1 o. I]

U
uh
LJllllnLlllr

 

 

 

 

 

 

-‘5 YTtTrfTTWI‘I'I'VUY'UY'VIIVII UIVKIIY‘YIUUTUIUUUIIIVIIIUIIII‘IIIIIIIYII'IIY -‘.5
4.5 -1 -0.5 0 0.5 1 1.5 -2 -1 0 1 2

Figure 5.7. Principle component analysis ordination of fungal T-RF LP peak area data
for microcosms containing two densities of A. albopictus (squares) and 0c. triseriatus
(circles), alone or under interspecific competition (triangles). High mosquito densities are
represented by filled characters and low densities by open characters. Panels A and B are

water column samples, C and D are leaf surface samples. Samples were taken on day 15

(A and C) and Day 30 (B and D). n = 3-4 for each point.

134

Discussion

In agreement with other studies, we have shown here that under conditions of
high larval density, competition for food resources in container habitats of larval
mosquitoes is intense and inherently changes the availability of such resources. For both
species, high densities negatively impacted larval success; however, interspecific
competitive interactions proved to be adverse only for DC. triseriatus populations. Larvae
of these mosquitoes feed primarily by grazing on the microbial biofilrn associated with
the container wall and detrital leaf surfaces or filtering microorganisms in the water
column; therefore, we postulated that measurable changes in the leaf and water column-
associated microbial community structure or abundance should occur at high interspecific
densities of two species. Our data do not support the hypothesis that A. albopictus
outcompete 0c. triseriatus by depleting resources below a level at which the latter can
survive if microbial abundance as a whole is the resource considered. We suggest instead
that the competitive advantage exhibited by A. albopictus may derive from a lower
resource-requirement in this species compared to CC. triseriatus due to their comparably
smaller body size.

Bacterial abundance and production rates did not differ among mosquito species
combinations although both measures were lower in high density treatments on day 15 in
response to higher grazing pressure. Bacterial abundance fell equally among the
treatments by day 30 and remained unchanged for the remainder of the experiment,
indicating that both mosquito species are depleting bacteria uniformly. Similar to the
present study, prior tests of density-dependent larval impacts on bacterial abundance have

not been supported (Kaufman et al. 2001), although decreases in leaf-associated bacterial

135

abundance and water column productivity occurred when larvae were present verses
absent.

Although not evaluated in the present study, the effect of mosquito feeding is
usually not associated with decreases in overall fungal abundance as indicated by
ergosterol concentrations in leaf material (Kaufman et a1. 2001, 2002). However, it
remains possible that differential harvesting of fungi or protozoans by the mosquito
species investigated here impacted competitive outcomes. Protists were not observed in
DAPI preps, indicating that this group was heavily grazed by both species and therefore
difficult to measure as a component of the competition mechanism.

Evaluation of terminal restriction fi'agments by principle component analysis
indicates that the native bacterial and fungal communities exhibit mosquito-induced
structural changes. Bacterial communities were influenced greatly by the density of
mosquitoes present in microcosms. This effect is unsurprising, given that under high
densities, feeding pressure increases and may in fact overwhelm the influence Of
individual mosquito species on community composition. Analysis of fatty acid methyl
ester (FAME) profiles has shown that larval feeding pressure changes the structure of
water column and leaf surface bacterial communities (Kaufman 1999). Subsequently, our
research group has constructed bacterial (16S) clone libraries to identify groups
associated with larval feeding (Kaufman et al. 2008, Xu et al. 2008). In particular, these
findings indicate that bacterial diversity decreases with larval feeding and the abundance
of certain groups, particularly the Flavobacteriaceae, increase in the absence of larvae

during mosquito feeding.

136

Despite a slight interaction of mosquito species composition with overall
mosquito density (significant at P < 0.1), the mosquito communities present in
microcosms had the greatest influence on the structure of fungal communities. Mosquito
species composition effected leaf —associated fungal communities on both day 15 and day
30. Previous reports have not indicated changes in overall fungal biomass, as estimated
by ergosterol concentration in response to larval feeding (Kaufinan et al. 2001, 2002),
although Kaufman et al. (2008) recently reported that several leaf-associated fungal taxa
exhibit shifts in response to larval feeding. When 00. triseriatus were present in tree
holes, increases in the abundance of Dothideomycetes and Saccharomycetes occurred
concurrent with decreased in the abundance of the Chytridomycota.

We did not observe compositional differences in microbial communities on day
60 associated with any of our treatments likely due to cessation of the majority of feeding
weeks earlier. Indeed, mosquito-associated changes in bacterial communities began to
dampen by day 30 of the experiment. Pupation of mosquitoes ended by day 34 and
although a small number of larvae remained in microcosms until day 60, their feeding
was insufficient for pupation and thus unlikely to contribute to significant shifts in
microbial community structure.

Of additional interest is the apparent temporal difference between bacterial and
fungal in community changes. While changes in the bacterial community were less
intense by day 30, changes in the fungal communities began to ramp up between day 15
and day 30, reflecting more significant influences of mosquito community and density on
fungal structure late in the mosquito development. This suggests that mosquito feeding is

having the greatest impact on fungi between days 15 and 30, unsurprising given that the

137

majority of mosquito adults emerged between day 15 and 25. The association of such a
shift with earlier development and greater emergence of A. albopictus under interspecific
competition with DC. triseriatus suggests that the former mosquito species is more
efficient at exploiting available firngal resources than 0c. triseriatus. Such differentially
utilization of fungal resources may underlie the resulting differences in mosquito
production observed when the species compete (Livdahl and Willey 1991, Novak et al.
1993, Teng and Apperson 2000). Resource-based competition has been described
between A. albopictus and 0c. triseriatus using leaves as the available resource
mediating competition. The role of fungi in mosquito production may be critical,
allowing mosquitoes to access nutrients otherwise available in the leaf matrix as they are
incorporated into fungal tissue (Kaufman and Walker 2006). Additionally, firngi would
be sources of essential lipids typically not present in bacteria. Many of these lipids are
necessary for emergence and flight (Dadd 1973), thus impacting directly on adult
production rates from the larval habitat. Although fungi support mosquito development
by increasing leaf decomposition, and thus the abundance of fine particulates and
dissolved substances in the water column (Pelz-Stelinski, Chapter 3 unpublished data,
Kaufman and Walker 2006), this study is the first to indicate that mosquitoes may
differentially utilize fungi resulting in production differences among species. Such a
disparity may be due to explicit differences in the preference of mosquitoes for particular
fungal taxa, resulting in the community composition changes observed here. There is
substantial evidence from studies of other insect detritivores that fungal composition
influences feeding rates and that, conversely, feeding can influence fungal community

composition (Graca 2001, Rossi, 1985).

138

More likely, however, is a difference in food acquisition, utilization, or a
combination thereof inherent to the competing larvae. Within the context of our original
postulates, we suggest that one or more of the following mechanisms are responsible for
the result of significant interspecific competition among larvae: 1) species exhibit a
differential ability to harvest and/or metabolize available fungal forms, 2) species exhibit
differences in their overall feeding behavior such that microbes are harvested
differentially from resource reservoirs i.e. water column verses leaf surfaces, and 3) A.
albopictus may drive down the total available fungal resources to a level below which the
competing species cannot maintain comparable productivity.

Understanding the mechanism driving competitive advantages is of particular
importance when one species is capable of competitively displacing the other, as in the
case of A. aegypti, displaced by A. albopictus in many habitats located in the southern
United States (see in Lounibos et al. 2001, Juliano and Lounibos 2005). In situations of
invading species, displacement of native mosquito populations may carry the risk
increased dissemination of disease if the invasive species is a superior bridge vector. This
is true of A. albopictus and 0c. triseriatus, where although the latter has the potential to
serve as a bridge vector, its preference for feeding on mammalian hosts makes Oc.
triseriatus a less competent vector of diseases such as West Nile Virus than the more
opportunistic feeding of A. albopictus (Turell et al. 2005). Livdahl and Willey (1991)
indicate displacement of 0c. triseriatus in man-made containers by A. albopictus is likely
where the two co-occur. Because the range of A. albopictus does not extend to Michigan,
0c. triseriatus has remained relatively free fi'om competition with the occasional

exception of native Culex spp. This interaction has not, however, resulted in the

139

displacement of 0c. triseriatus from their habitat. In contrast, 0c. japom'cus larvae have
been concurrently found with 0c. triseriatus in Michigan during 2005 and 2006 (personal
Observation). If it were to occur, competitive displacement of 0c. triseriatus poses a
threat to human health because 0c. japonicus is a superior vector of West Nile Virus and
is the only mosquito in Michigan with the capacity to vector Japanese encephalitis (U. S.
Centers for Disease Control and Prevention 2006). Evaluations of differential resource
usage by container-dwelling mosquitoes under interspecific and intraspecific conditions
may therefore provide not only a novel example of a complex trophic interaction, but
would also open the possibility for a practical means of relieving displacement through

the manipulation of specific microbial groups.

140

CHAPTER 6

INTERSPECIFIC MICROBIAL RESOURCE UTILIZATION IN AEDES ALBOPIC T US

AND AEDES AE GYPT I

141

Introduction

Container habitats are complex communities containing a diversity of
macroinvertebrate and microbial fauna. The dynamics of macroinvertebrate interactions
have been the focus of a large body of research, particularly due to the association of
such habitats with mosquito species. Nearly 40% of mosquito species utilize container
habitats, which include tires, tree holes, and commentary vases (Laird 1988). Among
these species are many important vectors of arboviruses, such as LaCrosse encephalitis,
yellow fever, dengue, and West Nile Virus. The food webs within container habitats are
generally driven by a variable milieu of allochthonous inputs, characterized by leaf and
animal detritus, and stemflow (Kitching 2001). Historically, the contribution of such
inputs has served as the basis for understanding the magnitude of mosquito production
from containers. Previous studies have indicated that the quantity and quality of detrital
resources, particularly leaf material, are of utmost importance to mosquito development
(Lounibos et al. 1992, Walker et al. 1997). It has become increasingly clear, however,
that the micrbiota supported by allochthonous inputs play an intermediate role in
container food webs, ultimately driving mosquito production.

The community dynamics of microorganisms associated with the water column
and detrital leaf surfaces have been assessed predominately via measurements of bacterial
abundance and productivity and fungal biomass (Kaufman et al. 1999, 2001, 2002, 2006),
although more recently bacterial and fungal taxonomic shifts have been identified t 16S
and 18S rRNA gene sequence analyses (Kaufman et al. 2008, Xu et al. 2008). These

studies demonstrated that larval feeding reduces the abundance and productivity of

142

bacterial leaf surfaces, while also changing the structure of both fungal and bacterial
communities associated with leaf surfaces.

The response of microbial communities to the presence of mosquito larvae has
been limited to studies of the Eastern tree hole mosquito, Ochlerotatus triseriatus (Say).
Although 0c. triseriatus often occur as intraspecific populations in the Northern portion
of their range, more often multiple mosquito species co-occur as a result of their
overlapping ranges and container preferences. Such is the case with two important vector
species, Aedes albopictus and Aedes aegypti. The competitive interactions underlying the
distribution of these species have been the subject of much interest given the capacity of
both species for exerting conditional displacement of the other under interspecific
competition. Several studies have shown that declines in North American populations of
A. aegypti correlate to the appearance of A. albopictus (O'Meara et al. 1993, 1995, Hobbs
et al. 1991, Homby et al. 1994, Mekuria and Hyatt 1995). In contrast to urban habitats
wherein co-occurrence may take place, displacement of A. aegypti frequently occurs in
suburban or rural habitats (Homby et al. 1994, O'Meara et al. 1995). Juliano (1998)
showed that the decline of A. aegypti is probably driven by interspecific resource
competition among larvae. Under resource limited conditions, A. albopictus develop
faster and attain a greater body size that do A. aegypti when the species co-occur (Barrera
1996, J uliano 2004); however, this effect was mitigated when interspecific populations
were provided with abundant resources (Daugherty et al. 2000).

Apparent competition has been suggested as a mechanism driving the competitive
outcomes and may occur via characteristics exhibited during non-competing life stages.

For example, condition-specific responses to fluctuating abiotic conditions are

143

responsible for the outcome of competition between Aedes aeqypti and A. albopictus. The
latter exhibits a competitive edge over A. aegypti under wet (normal) conditions,
sustaining greater population growth at low resource levels (Costanzo et al. 2005). Under
dry environmental conditions, however, the competitive advantage shifts to A. aegypti
due to their comparatively higher tolerance of eggs to desiccation.

Differences in modes of feeding between the two species may be responsible for
the competitive advantage exhibited by A. albopictus. Although we do not attempt to
quantify such differences here, relative differences in amount of filter feeding versus
surface grazing should be reflected in the microbial communities associated with these
substrates under interspecific and intraspecific conditions. In addition, the capacity of A.
albopictus to reduce populations of overall microbial abundance, or the abundance of
individual microbial taxa, below a level that will sustain A. aegypti may also contribute to
the outcome of interspecific competition.

In this paper, we describe the response of microbial communities to the presence
of A. albopictus and A. aegypti alone or under interspecific competition. Specifically, the
objectives were to determine: 1) whether shifts in the structure of microbial communities
occur in response to overall mosquito density? or 2) if the composition of mosquito
communities in containers has a specific impact on the dynamics of microorganisms that
contributes to the displacement of A. albopictus by A. aegypti? If general resource
abundance is responsible for the success of A. albopictus, we predict that microbial
community composition would not respond differently when the species compete,

whereas we predict that if larval competition is dependent on the structure of microbial

144

communities, changes in the relative abundance of microbial community members should
be apparent under interspecific verses intraspecific conditions.

Materials and Methods
Microcosms. We investigated the microbial dynamics underlying interspecific
competition between A. albopictus and A. aegyptg' in laboratory microcosms simulating
natural container habitats. To construct microcosms, senescent red oak leaves (Quercus
rubra L.) were obtained from Michigan State University’s Kellogg Forest (Augusta, MI),
dried at 45°C for 48 h, and added to containers as 1.0 g leaf packs. Three days prior to the
addition of mosquitoes, 500 ml deionized water and a 3 ml of microbial inoculums were
also added to the containers. The microbial inoculums consisted of composite water and
particulates randomly sampled from tree holes located in woodlots near the Michigan
State University campus (East Lansing, MI) and homogenized in a standard kitchen
blender. After allowing microcosm microbiota to condition for three days, newly hatched
first instar A. albopictus and A. aegypti were added to microcosms (day). Larvae were
procured from our laboratory colony, originally obtained fi'om the Malaria Research and
Reference Reagent Resource Center (Manassas, VA). Larvae were added in total
densities of 30 or 60 total larvae with the following rations of A. aegypti: A. albopictus:
0:30, 30:0, 15:15, 30:30, 0:60, and 60:0. In addition, no larvae microcosms were included
to assess the overall affect of mosquitoes on microbial community dynamics. Twelve
replicates were constructed per treatment combination, to allow for destructive sampling
of six replicate microcosms on day 7 and day and day 14. These dates were chosen to

obtain measurements of microbial activity during the time the majority of larval

145

development occurred. Throughout the experiment, microcosms were held. in an
environmental chamber at 28° C and 16L: 8D.

Sampling. Microcosms were monitored daily for mosquito adults. After eclosing, adults
were collected, sexed and identified before undergoing liophilization for subsequent dry
mass measurement. Larvae remaining in microcosms upon destructive sampling were
collected and similarly identified, lyophilized and massed; identifications to species were
not made below the third instar, nor were larvae distinguished by sex, due to the
difficulty of observing relevant morphological characters. In addition to mosquitoes,
samples of leaf material and microcosm water were taken on day 10 and day 17for
analysis of microbial community structure and bacterial productivity (described below).
Leaf samples were aseptically obtained using a sterilized cork borer (1 1 mm), placed in
filter-sterized phosphate-buffered saline, and sonicated on ice for 12 min. to dislodge the
loosely-attached fraction of microorganisms. Leaf sonicates were utilized rather than
whole-leaf extracts because this fraction represents microorganisms physically available
to grazing mosquito larvae.

Bacterial production. Accumulation of bacterial biomass in water column and leaf
sonicate microcosm samples was directly measured via quantifications of 3’H-leucine
incorporation (Kirchman 2001). Using bacteria-specific nanomolar leucine
concentrations of 25 nM (Smith and Azam 1992), we measured the incorporation of the
labeled amino acid following incubations of the samples with 3H-leucine (L-leucine (4,5-
3H), 50 CI/mmol- NEN, Life Science, Boston, MA). Unlabeled leucine was added with
labeled leucine, bringing the total leucine concentration in leaf and water samples to 400

and 100 nM, respectively, to account for saturation of uptake kinetics (Kirchman 2001,

146

Kaufman et al. 2001). Following incubation of samples at 20°C for 20 min,
trichloroacetate (TCA; 10% volzvol, final concentration) was added to precipitate protein
and terminate incorporation of additional leucine. Two additional TCA rinses and a
single ethanol (80%) rinse were used to precipitate additional protein. Finally, samples
were suspended in scintillation cocktail for quantification of radioactivity with a
Beckman LS6500 scintillation counter (Beckman Coulter, Inc., Fullerton,CA).
Microbial community analysis. T-RFLPs. The structure of leaf- and water column-
associated bacterial and fungal communities in microcosms was assessed using tagged
restriction fragment length polymorphism (T-RF LP) analysis (Marsh 1999). DNA was
extracted from the samples described above with the MoBio UltraClean soil extraction kit
(Carlsbad, CA). After ensuring the success of extractions via PCR and agarose gel
electrophoresis, sample DNA was amplified using labeled (5’-FAM) fungal (l8S rDNA)
(Bomeman and Hartin 2000) and bacterial (16S rDNA) (Marchesi et al. 1998) primers in
independent PCR reactions. The primers used for fungal and bacterial amplifications
were, respectively: nu-SSU-0817-59F (5’-TTAGCATGGATAATRRAATAGGA-3’),
nu-SSU-1536-39R (5’- TTGCAATGCYCTATCCCCA-3’); and 63F (5’-
CAGGCCTAACACATGCAAGTC-3 ’), 1387R (5’-GGGCGGWGTGTACAAGGC-3’)
(Integrated DNA Technologies, Coralville, IA). Co-purifications of products from three
100 111 PCR reactions per primer set were conducted using the Qiaquick PCR purification
Kit (Valencia, CA). Subsequently, PCR products were subjected to digestion with the
restriction enzyme, MspI (New England Biolabs, Cambridge, MA) according to the
methods described by Marsh (1999). Samples were then submitted to the Michigan State

University Research and Technology Support Facility for electrophoresis of digested

147

samples on a capillary electrophoresis genetic analyzer (PRISM 3100 Genetic Analyzer,
Applied Biosystems, Foster City, CA). Size and frequencies determinations for each
terminal restriction fragment were determined using Genescan Analysis software (version
3.7, Applied Biosystems). Only samples with combined peak areas of 10,000 or greater
were retained for subsequent analyses.
Statistical Analysis. T-RF LP data from 16S and 18S rRNA gene communities were
subjected to principal component analysis (PCA) using R (R Development Core Team
2004). The principal components (PCs) , derived from the linear combination of peak
area abundances for individual fragments, were analyzed by standard analysis of variance
(AN OVA) to evaluate the main effects of mosquito species (interspecific or intraspecific
populations), sampling date (day 7 or day 14), and larval density (0, 30, or 60). Mosquito
production values (mass and development time for males and females) and bacterial
productivity (per microcosm) were also analyzed by AN OVA, followed by Bonferroni
adjustment of experiment error from a = 0.05. Following the ANOVA, Tukey’s HSD
post hoc test was used for separation of treatment means. Main effects contributing to
mosquito production were limited to species and density, as only mosquitoes obtained
from microcosms retained until the end of the experiment (day 24) were included in the
analysis.

Results
Mosquito production. For both mosquito species, male adult mass was significantly
affected by mosquito treatment (Table 6.1); however, in both cases male mass responded
to the overall effect of density rather that species composition, as mosquitoes at low

density (30 larvae per microcosm) attained greater average body mass that mosquitoes in

148

high density treatments (60 larvae per microcosm) (Figure l and 2). A similar trend is
suggested for female mosquitoes of both species, although only changes in A. aegypti
adult female body mass were significant following Bonferroni correction (Table 6.1,
Figure 6.2). At high total mosquito densities (alone or with A. albopz'ctus), females body
mass was significantly reduced compared with low density treatments in that only
microcosms containing low mosquito densities produced adult females.

The time for development of A. albopictus from larvae to adults was not
significantly impacted by mosquito species, and this effect was consistent between males
and females (Table 6.1, Figure 6.1). In contrast, male and female A. aegypti development
time was significantly affected by the absence of adult emergence under high density
treatments, when the species was alone or under interspecific competition (Table 6.1,

Figure 6.2).

149

Has—m 97 >ZO<> 89:8 m3 80358 30330: <3%—am.

 

 

 

 

”Became <m1mgo mofioo Rama 28325. Amuse mmwvfiz.
E. m. w a». m. w
7.3—a 38m muoomom 7No VB 0.0m 7B 73 9w
@3me 78 3.3 AoboS * 7B omd AoboS a.
258 923355.: muoomam 78 PS obn 7B 783 one
U035, 7mm 93 98 73 mm .3 AoboS *
mmaao mem muoommm 7mg Pmu o. H N 75 wk obo
63me 7.8 obs oboo... 78 3.3 92: ._.
mmBEo ago—839: muoomam 73 9mm 93 73 PE 9:
D953, 7mm 9d obg 73 3.3 983...

 

£59088 amamomboo mo=o§=m webmafloa 2:52:03. 588303 6328: $598 85 @3me :2 £52: 5 820 $58 so"

mwmamoma Aqume.

150

A. albopictus
A a

0.3

 

0.25
0.2 . l a

a
0.15 i .

0.1

1b 1b

Mass (mg)

 

 

 

 

16

 

I {

days

 

 

0 . . 1 . . .
A. albopictus 15 30 30 60 15 30 30 60
A. aegypti 15 0 30 0 1 5 0 30 0

 

 

 

Figure 6.1. Mosquito production variables for A. albopictus males (A and C) and females
(B and D). Average mosquito weight (A and B) and development time (C and D) are

shown as means :1: SE. n = 6 for variables in A and B, n = 2-6 for variables in C and D.

151

A. aegypti
0.3

 

 

0.25

0 a
0.15 {a

0.1

 

Mass (mg)

0.05

O
P
10'
00.
e
O'
U

 

 

 

 

12 {

days

 

 

b b
0 1 e e b b

A. aegypti 15 30 30 60 15 30 30 60
A. albopictus 15 0 30 0 15 0 3o 0

 

 

 

 

Figure 6.2. Mosquito production variables for A. aegypti males (A and C) and females (B
and D). Average mosquito weight (A and B) and development time (C and D) are shown
as means :1: SE. n = 6 for variables in A and B, n = 0-6 for variables in C and D.
Treatments means with different letters are significantly different following Bonferroni

correction for experiment wide error.

152

 

Bacterial productivity. Bacterial production on leaf surfaces responded
significantly to the main effect of mosquito species composition on day 7 of the
experiment; however, this effect was not apparent on day 14 (Table 6.2). In contrast, the
affect of mosquito species composition on bacterial production in the water column was
significant on day 14, but not day 7. The significance of species in this case likely a result
of the significant interaction of species and density on bacterial productivity, as
production of bacteria was remained high in microcosms without mosquitoes on day 14,
but was comparably lower when mosquito larvae were present. A similar significant
interaction occurred for bacterial production associated with leaf surfaces. For both
substrates (leaf and water column), mosquito density did not significantly affect the

production of bacteria on day 7 or day 14.

153

Table 6.2. AN OVA results for bacterial productivity.

 

 

 

Source Water Leaf
df F P (if F P
Day 7 Species 3 0.53 0.08 3 12.79 <0.0001*
Density 1 0.08 0.78 l 1.08 0.31
Species*Density 3 2.95 0.07 3 0.04 0.96
Error 32 32
Day 14
Species 3 8.41 0.0003* 3 1.52 0.22
Density 1 0.39 0.53 1 2.99 0.09
Species*Density 3 5.45 0008* 3 10.73 0.0002*
Error 34 34

 

*Indicates significance following Bonferroni adjustment. Species treatment refers to the

effect of mosquito species composition on bacterial production. Interactions between

species and density not shown in table were not significant (p<0.05).

154

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

16
Dday7 lday14 A
14 . "
12 - 4
.C 10 i '-
\
.8 8 .
'0
> 5 4
O
E 4 *
Q.
0 w . ~ T . f
6
B
.c 5 i
=
E 4 ~
6
E 3 “
o.
2 s
1 d fl
0 I I I I I i
A. albopictus 0 15 30 0 30 60 0
A. aegypti 0 15 0 30 30 0 60

Figure 6.3. Bacterial productivity in microcosms on day 7 and day 14 following larval

addition. (A) Leaf surface. (B) Water column. Values are means i SE (n=5-6).

155

T-RFLPs. Using PCA to derive new variables representing linear combinations of
individual microbial populations represented by T-RF LP fragments, we determined that
microbial communities responded significantly to the main effects in this experiment.
Bacterial communities on leaf surfaces changed significantly between days 15 and 30, as
indicated by the significant affect of time on PC’s 1 and 2 (Table 6.3, Figure 6.4). The
main effect of competition type combination did not significantly affect the structure of
bacterial communities along either PC axis; however, PC1 representing leaf-associated
bacterial communities was significantly affected by the density of mosquitoes in
microcosms (Table 6.3). Although bacteria in water column samples obtained from
microcosms were not sufficient to permit robust univariate analysis of treatment effects
on principal component axes, PCA revealed that the overall affect of larval presence was
sufficient to alter the structure of bacterial communities (Figure 6.5).

Of the two mosquito species, only A. aegypti presence had a significant affect on
leaf-associated fungal communities represented by PC1 (Table 6. 4, Figure 6.6). Density,
on the other had, exerted a significant affect on fungal communities represented by PC2
in microcosm containing both species, and by PC1 in microcosm containing A. aegypti.
In addition, the main effect of time on PC1 was significant in microcosms in which A.

albopictus appeared, although a similar significant effect was not apparent for A. aegypti.

156

Table 6.3. Summary of ANOVA results for principal component (PC) scores from leaf

surface bacterial (16S rRNA gene) T-RFLP profiles.

 

 

 

PC1 PC2

Source df F P F P
Competition type 2 8.05 0.006 1.98 O. 15
Density 1 1.04 0.36 0.71 0.40
Time 1 22.39 <0.0001 14.31 0.0004
Competition*Density 2 0.19 0. 83 1.1 1 0.34
Competition*Time 2 0.33 0.72 0.14 0.71
Density*Time 1 O. 14 0.71 0.42 0.66
Competition*Density*Time 2 0.41 0.67 1 .38 0.26

Error 59

 

157

Table 6.4. Summary of ANOVA results for principal component (PC) scores from leaf

surface fungal (188 rRNA gene) T-RFLP profiles.

 

 

 

PC1 PC2
Source df F P F P
Competition type 2 7.22 0.002 0.66 0.52
Density 1 5.97 0.02 17.24 0.0002
Time 1 0.00 0.98 12.93 0.001
Competition*Density 2 5 .65 0.008 O. 16 0.85
Competition*Time 2 0.92 0.41 4.29 0.02
Density*Time 1 0.64 0.43 0 0.96
Competition*Density*Time 2 0.88 0.36 5.52 0.02
Error 59

 

158

 

 

 

 

0.1 a +_ __,_._ r
l 0 Day 7_o
I a Day 7_3o
0.05 w '
1 a Day 7_60
S3 I I i 0 Day 14_0
l0 .
3i 0 l D Day 14_30 |
R? + I a Day 14_60
-o.05 ,
-o.1
'O.15 . I I I

-O.15 -O.1 -0.05 O 0.05 0.1 0.15

PC 1 (25.6%)

Figure 6.4. PCA of TRF LP fragment peak areas from leaf surface bacterial (16S rRNA
gene) communities. Total density of mosquitoes (O, 30, or 60) are compared for each

sampling date (Means :1: SE). The two component axes explain 48.1% of the variation.

159

 

1.2

 

 

 

 

 

7:636“
1 '——‘—| I 15:5
|
°\° A 30:0 1
g ' I D 30:30 ‘
8 0.4 i 0 0:60 i
0‘ I 0 60:0
0.2 -. ’ “r i “J
O P f 1
-0.2
-0.1 -0.05 0 0.05 0.1 0.15

PC 1 (35.1%)

Figure 6.5. PCA of TRFLP fragment peak areas from water column bacterial (16S rRNA
gene) communities. Individual A. albopictuszA. aegypti treatments are compared (Means

i SE). The two component axes explain 60.6% of the variation.

160

 

 

 

1.36—1
6 ._ -15:15i
I——O—-I
4 A0230
g I—I—d A30:0
5.5 2 I—a-I D 30:30
8 0 00:60
‘1 99939-1
-2
-4 - ' i ' H—é—H
’6 I . I l

 

 

 

-1.5 -1 D5 0 0.5 1 1.5 2
PC1(17.5%)

Figure 6.6. PCA of T-RF LP fragment peak areas from leaf surface fimgal (18S rRNA
gene) communities. Comparisons of individual A. albopictus: A. aegypti treatments are

shown (Means :1: SE). The two component axes explain 27.9% of the variation.

161

Discussion

Performance of mosquito populations in microcosms was consistent with findings
from several key studies which defined the nature of competition between A. albopictus
and A. aegypti (Barrera 1996, J uliano 1998). Although both species fared poorly under
high density and hence, low resource treatments, adult A. albopictus production
continued through the experiment. Male and female A. aegypti, in contrast, failed to
emerge from each high density treatment. Thus, A. aegypti was more susceptible to
negative effects conferred by high densities than A. albopictus.

Interestingly, despite prior studies to the same effect in nature, this is the first
study to describe A. albopictus success over A. aegypti in a microcosm environment.
Previous microcosm studies found A. aegypti to be a superior competitor in laboratory
studies (Moore and Fisher 1969, Ho et al. 1989, Black et al. 1989); however, in these
experiment, artificial diets were employed rather that leaf material microorganisms as the
resource base. As in the current experiment, Juliano (1998) utilized plant detritus in field
experiments evaluating competitive outcomes between the two species. Finding that A.
albopictus was competitively superior to A. aeqypti at high densities, he postulated that
the different resources used in these studies promoted differential outcomes when the two
species compete. Hence, clearly defining the resource utilized by competing mosquitoes
is essential to understanding competitive outcomes. While the important of resource type
has been illustrated by the above cases, subsequent studies have not addressed mosquito-
mediated changes at the level of microorganisms. If we consider that mosquitoes

consume microorganisms as a food resource in order to access the nutritional reserves

162

provided by plant detritus, then we would expect that mosquito feeding may affect
microbial communities differentially in a species-specific manner.

The microcosms in this study were designed to describe the dynamics underlying
bacterial and fungal communities associated with mosquito species competition.
Although the use of 16s rRNA gene sequence data has been used previously to describe
microbial populations, the technique is of limited use in replicated experimental designs
for economic and practical reasons. The T-RFLP technique, by comparison, permits a
finer level of taxonomic resolution for a large number of samples, wherein each fragment
defined can be construed as a unique “species.” The results obtained from principal
component analysis of variables derived from the relative abundance of microbial rRNA
gene fragments, it is apparent that mosquito species indeed change the constitution of
bacterial and fungal communities associated with leaf material and, in the case of
bacterial communities, the water column.

Regardless of the mosquito species composition represented in microcosms (e. g.
interspecific or intraspecific populations), density appeared to be the most important
factor contributing to variation in bacterial communities. Both leaf-associated and fimgal
communities exhibited changes that are undoubtedly related to the effects of mosquitoes
grazing down populations of microorganisms. Indeed, that a time effect was evident from
day 7 to day 14 is consistent with the hypothesis that the microbial community decreases
over time in the presence of larvae. That mosquito presence affects microorganisms has
been clearly determined by previous microcosm studies which report not only reduction
of biomass but also shifts in community composition (Kaufman et a1. 1999, 2001, 2002,

2006, 2008).

163

In the current study, we have sought to determine whether mosquito feeding alters
microbial communities in a species-specific manner. We found that the composition
mosquitoes present only affected leaf-associated bacterial communities when
intraspecific A. aegypti populations were compared with interspecific mosquito
communities. Although a marginal effect of species composition was also found in
comparisons of intraspecific and intraspecific A. albopictus communities, the effect was
not significant following a Bonferroni adjustment for multiple comparisons. Thus, the
shift in fungal structure only when A. albopictus are present suggests that they may
differentially altering the fungal community on leaf surfaces compared with A. aegypti.
The mechanism for this effect is unclear, although we postulate that it stems fi'om
differences in the development of the species under larval competition. Kaufman et al.
(2008) suggested that 0c. triseriatus larvae at multiple stages of development may
differentially alter fungal communities in loosely attached verses adherent fractions of
leaf material. Also, in stream communities, use of diatom communities by different
macroinvertebrate groups was in part a function of consumer size (Tall et al. 2006).
Because overall mosquito feeding contributes to changes in microbial communities, we
suspect that a faster developing species, such as A. albopictus, should affect such changes
more rapidly than A. aegypti, resulting in the difference fungal communities when A.
albopictus are present. A similar result would be expected in intraspecific mosquito
populations as well if a comparison was made between a population comprised of a
single cohort (that is, the same instar) and a mixed population of early and late instar

larvae.

164

In the past, our research group has demonstrated that mosquito feeding impacts
the microbial community in tree holes. The changes wrought by larval activity have
manifest as decreases in the abundance of bacterial and fimgal biomass, and bacterial
growth rates associated with leaf material, although changes in firngal biomass have been
less evident (21-24). In addition to these “big picture effects,” changes in the structure of
microbial communities have also been evident in response to larval feeding. Shifts in
major taxonomic groups are reflected in both bacterial and fimgal communities
associated with leaf material in tree holes. Evidence from 16S rRNA gene sequence
analysis suggests that larval presence selectively effects members of the
Betaproteobacteria, Alphabroteobacteria, and Bacteriodetes bacterial classes, resulting in
depression of abundance in the latter two groups and enhancement in the former
(Kaufinan et al.2008). Similarly, Dothideomycetes and Saccharomycete fungal taxa
decreased in response to larval feeding, while the abundance of Chytridiomycota
members increased. Recently, we have utilized T-RFLP analysis to evaluate larval
feeding effects on microbial communities in a fashion that permits both replication and
greater resolution of microbial taxa (Pelz-Stelinski, Chapter 2). Compared with the
results obtained from sequence analysis, T-RFLP results reflected a considerable amount
of variation among replicates such that treatment differences were difficult to resolve.
Indeed, variation in among treatments is also characteristic of sequence analysis carried
out at fine resolution (Pelz-Stelinski, Chapter 2 and Kaufinan et al. 2008); however, as
the taxonomic level of resolution chosen for community analysis becomes increasingly
broad (i.e. genus level verses class level), larval feeding effects become apparent. This

may suggest that the level of resolution provided by T-RFLPs, that is to say “species-

165

level” (Marsh 1999), may not be biologically relevant from the mosquito perspective.
The current findings suggest that the effect of combined mosquito density drives shifts in
the structure of microbial communities rather than individual species. We postulate that,
in addition to other possible factors regulating resource-based competition between the
two species, either or both of the following mechanisms might also be possible: 1)
microbial groups shift in response to mosquito feeding, but at a higher level of taxonomic
resolution not evident here; or 2) mosquito species differ in their physiological or
behavioral capacity for assimilating/harvesting the nutritional resources provided by

microorganisms.

166

GENERAL CONCLUSIONS

167

In a series of experiments designed to evaluate the effects of detrital leaching on 0c.
triseriatus productivity, I determined that the soluble fraction of leaf detritus initially
released into container habitats is significantly associated with higher bacterial
production and greater mosquito success relative to containers containing leaves from
which the soluble component was removed. In addition, mosquito performance was less
robust when the leached fraction of leaves was provided in the absence of leaves,
although mosquito productivity in response to this treatment was still better than when
leached leaves alone were provided for developing larvae.

Several studies in this dissertation employed terminal restriction fragment
polymorphism (T-RFLP) analysis to obtain profiles of the fungal and bacterial
communities associated with container habitats. Comparisons of field and laboratory
container habitats showed that, in addition to temporal fluctuations, the composition of
microbial communities change in response container type and mosquito density. Shifts in
the structure and dynamics of container—associated microbial communities were also
evident in response to macroinvertebrate species composition. Although the composition
of fungal communities exhibited a significant shift only in response to 0c. triseriatus
density, more firngal decomposition enzyme activity was associated with presence of
scirtid beetle larvae, which are known to facilitate the production of mosquitoes from tree
holes. Data from terminal restriction fragment polymorphism (T-RF LP) analysis
suggested that scirtid presence influenced bacterial communities associated with leaf
material and the water column. Furthermore, I showed increased processing of leaf
detritus, higher leaf-associated enzyme activity, and higher leaf-associated fungal

biomass was due to scirtid presence. Such shifts suggest beetle feeding facilitates

168

mosquito production indirectly through the microbial community rather than directly
through an increase in available fine particulate organic matter (FPOM).

Different species often utilize similar available food resources, but show
differential growth under competitive conditions. Using experimental microcosms, I
examined the effect on microbial community dynamics and community structure in
response to competition between A. albopictus and 0c. triseriatus, and between A.
albopictus and A. aegypti. I postulated that leaf- and water column-associated microbial
community structure and abundance would differ in microcosms containing interspecific
mosquito populations compared with intraspecific populations. Patterns of T-RFLPs
obtained from microcosm studies suggested differences in the diversity microbial
community under conditions of interspecific versus intraspecific competition. The
hypothesis that mosquito species exhibit a differential ability to survive at low microbial-
resource levels and/or to digest the microorganisms available was supported by changes
in the leaf and water column-associated microbial community structure, productivity, and
abundance under interspecific competition between A. albopictus and 0c. triseriatus and
between A. albopictus and A. aegypti.

The studies described in this dissertation cumulatively illustrate changes in the
structure of fungal communities associated with mosquito presence. Shifts in the fungal
community were evident in response to mosquito species composition and overall
mosquito density. Although the importance of these shifts to developing mosquitoes has
not been determined, it is clear that In addition, these studies illustrate the dependence of
microbial dynamics on the initial microbial inoculum present in container habitats.

Differences among container habitat types varied between field and laboratory

169

experiments, suggesting the need for further research into the relevance of particular
microbial assemblages to mosquito development. In addition, future work is should
assess the response of protozoan communities to mosquito presence. Although not a
component of these studies, it is likely that protozoans groups will also exhibited shifts
under the regimes evaluated in this dissertation, as they are similar to bacterial and fungal

communities in their role as a nutritional resource for developing mosquito larvae.

170

APPENDIX 1.

RECORD OF DEPOSITION OF VOUCHER SPECIMENS

171

Appendix 1.1

Record of Deposition of Voucher Specimens*
The specimens listed on the following sheet(s) have been deposited in the named
museum(s) as samples of those species or other taxa, which were used in this research.
Voucher recognition labels bearing the Voucher No. have been attached or included in
fluid-preserved specimens.
Voucher No.: 2008-05
Title of thesis or dissertation (or other research projects):

MOSQUITO PRODUCTION AND MICROBIAL DIVERSITY IN CONTAINER
HABITATS

Museum(s) where deposited and abbreviations for table on following sheets:

Entomology Museum, Michigan State University (MSU)
Other Museums:

Investigator’s Name(s) (typed)
Kirsten Pelz-Steliniki

 

 

 

Date April A 2008

*Reference: Yoshimoto, C. M. 1978. Voucher Specimens for Entomology in North
America.
Bull. Entomol. Soc. Amer. 24: 141-42.

Deposit as follows:
Original: Include as Appendix 1 in ribbon copy of thesis or dissertation.

Copies: Include as Appendix 1 in copies of thesis or dissertation.
Museum(s) files.

Research project files.

This form is available from and the Voucher No. is assigned by the Curator, Michigan
State University Entomology Museum.

172

Appendix 1.2

Voucher Specimen Data

Page_1_of A Pages

 

am was: as a €88

 

Marmara 23

 

 

 

 

8.: mangoomm can: o>o:a o5 3383:

 

8-88 .02 .8:uao>

38:05.50: 8:5.

Avonbv A3252 {Sawumozd

A533: a: $021 Quota? omDv

 

 

 

 

 

 

 

 

 

 

 

 

 

 

E28 :2 3&5 am: 88 632 am

DmE o: waded“: «mam do an: 53:32 $3355 :QAMNE .8an
328 a: 39:8 2m: 88 532 am

2m: S 9:83 Ham .8 5&5 5322 aaoaaad .388 38w
328 a: 8980 am: 88 £32 3.

3m: 0: mam—8‘— .3m .00 83%.: 53:22 $385.6 animus “show
328 a: 3950 am: 88 £32 a

pm: 2 mamas Ham .8 5&5 8322 323% 338% Ear
328 :fl 3950 Dm—Z Bow :83): 3

3m: S 383 cam .oo 8235 5322 9.3% 3.38% 88w
328 a: 8950 am: 82 £32 am

392 c: mamas: 6am .00 5&5 fiwEozz 33:9 3636:? .8an
mg 8:18an 88 ugfioumom 2

DE): 3 Bufimmm do Bufiwam gwfloaz 8.38:5 gonzomuanSEfifiQ

d ,6 0+

m m .m r m m e .m. R 3:83: :85 8:8 8 86on

m. e m. e .m u m. m. m % :5 :8: no “388:8 8258mm 3.: 8% 6:3
h e w d
M w d A A W N L E
«o 5:852

173

Appendix 1.2

Voucher Specimen Data

ges

e20f_2_Pa

Pag

Ban 8350

 

.8332 Eofifiofim

 

8.572 28

 

basins 83m gwfioaz o5 E :89:

 

8.: 3250on was: o>o:a 05 330qu
3-38 .02 6:25>

 

Eéoamfiom 5qu
825v 30:82 9:083:82:
38380: t 83% 3quan 83

 

328 a: 8950 am: 88 £32 a
DmE c: wanna: «mam do 85:»:— gwfloaz
€28 :2 3:50 am: 88 632 am

Axmmv «Eureka: wESEmEuQ

 

 

 

 

 

 

 

 

 

 

 

 

 

DE): 0: wfimg «mam do as gwfioaz Amamv 35.:me 358350
328 :2 3988 392 boom :33: am
Dmv: o— wfimqa: «mam .00 ngfl 5&322 955 33.233 3895390
10 0+ 5
m m .m r. m m e .m. m @8303: no 5 o S 883
m m m. M .w .m m. m. m m.» 28 com: 8 @3258 8230on 8a 38 3:3 x8 5 . m
M w d d A A m N L E
mo 3:852

174

APPENDIX 2.

SUMMARY OF ANOVA RESULTS FOR PRINCIPAL COMPONENT ANALYSIS OF

MICROBIAL COMMUNITIES ASSOCIATED WITH MICROCOSMS IN CHAPTER

3.

175

Appendix 2, Table 1. Summary of ANOVA results for PC1 scores from water column

16S rDNA T-RFs digests (Block I).

Source PC1 PC2
df F Value P Value F Value P Value

 

 

Msp
Mosquito 1 5.86 0.036 1.73 0.218
Scirtid 1 0.1 1 0.741 0.27 0.613
Time 1 21.06 0.001* 0.23 0.643
Mosquito x Time 1 2.58 0.139 0.33 0.577
Scirtid*Time 1 0.12 0.735 0.21 0.654
Mosquito*Scirtid 1 0.0 0.954 0.02 0.899
Error 10

Hha
Mosquito 1 1.05 0.328 0.04 0.838
Scirtid 1 10.69 0008* 1.35 0.270
Time 1 0.01 0.941 0.01 0.930
Mosquito*Time 1 0.21 0.656 0.00 0.974
Scirtid*Time 1 0.05 0.827 1.68 0.222
Mosquito*Scirtid 1 0.08 0.786 12.94 0004*
Error 11

 

*Indicates p-values that are significant following Bonferroni adjustment

176

Appendix 2, Table 2. Summary of ANOVA results for PC1 scores from water column

16S rDNA t-RFs digests (Block H).

—
Source PC1 PC2

df F Value P Value F Value P Value

 

 

Msp
Mosquito 1 6.75 0.019 53.23 <0.001*
Scirtid 1 16.93 <0.001* 1.94 0.184
Time 1 14.03 0.002* 27.9 <0.001*
Mosquito x Time 1 0.50 0.489 0.42 0.529
Scirtid*Time 1 7.37 0.015 0.05 0.824
Mosquito*Scirtid 1 1.00 0.333 0.27 0.614
Error 15

Hha
Mosquito 1 0.90 0.358 8.67 0.01*
Scirtid 1 1.74 0.207 0.32 0.58
Time 1 0.00 0.978 0.00 0.966
Mosquito*Time 1 0.00 0.970 0.34 0.571
Scirtid*Time 1 0.56 0.465 0.54 0.473
Mosquito*Scirtid 1 0.95 0.354 0.62 0.444
Error 15

 

*Indicates p-values that are significant following Bonferroni adjustment

177

Appendix 2, Table 3. Summary of ANOVA results for PC1 and PC2 scores from leaf

surface 16S rDNA t-RFs digests (Block I)

_
Source PC1 PC2

df F Value P Value F Value P Value

 

 

Msp
Mosquito 1 2.94 0.106 0.51 0.486
Scirtid 1 1.92 0.185 0.02 0.895
Time 1 1.90 0.187 4.08 0.060
Mosquito x Time 1 1.65 0.217 0.41 0.532
Scirtid*Time 1 0.39 0.539 0.00 0.344
Mosquito*Scirtid 1 1.61 0.223 0.95 0.810
Error 16

Hha
Mosquito 1 3.23 0.094 0.43 0.521
Scirtid 1 0.61 0.450 1.47 0.245
Time 1 0.72 0.410 3.22 0.094
Mosquito*Time 1 0.02 0.896 0.00 0.953
Scirtid*Time 1 0.00 0.971 6.63 0.022
Mosquito*Scirtid 1 0.74 0.736 0.34 0.569
Error 14

 

178

Appendix 2, Table 4. Summary of ANOVA results for PC1 and PC2 scores from leaf

surface 16S rDNA t-RFs digests (Block II).

—
Source PC1 PC2

df F Value P Value F Value P Value

 

 

Msp
Mosquito 1 41.17 <0.001* 2.16 0.162
Scirtid 1 5.33 0.036 0.19 0.672
Time 1 8.04 0.012* 2.0 0.178
Mosquito x Time 1 1.39 0.257 0.60 0.449
Scirtid*Time 1 0.10 0.756 0.85 0.372
Mosquito*Scirtid 1 3.1 1 0.098 0.10 0.101
Error 15

Hha
Mosquito 1 1.85 0.194 3.12 0.098
Scirtid 1 5.09 0.039 3.10 0.099
Time 1 0.20 0.661 0.03 0.868
Mosquito*Time 1 3 .03 0.102 0.54 0.475
Scirtid*Time 1 0.03 0.858 0.31 0.589
Mosquito*Scirtid 1 1.80 0.200 0.40 0.539
Error 15

 

*Indicates p-values that are significant following Bonferroni adjustment

179

Appendix 2, Table 5. Summary of ANOVA results for PC1 and PC2 scores from water

column 188 rDNA t-RFs digests (Block I).

—
Source PC1 PC2

df F Value P Value F Value P Value

 

Hha
Mosquito 1 10.02 0.016* 0.15 0.706
Scirtid 1 1.72 0.231 0.30 0.601
Time 1 1.45 0.267 8.05 0.025
Mosquito*Time 1 0.96 0.360 0.53 0.489
Scirtid*Time 1
Mosquito*Scirtid 1 0.02 0.900 0.03 0.873
Error 7

 

*Indicates p-values that are significant following Bonferroni adjustment

180

Appendix 2, Table 6. Summary of ANOVA results for PC1 and PC2 scores from water

column 188 rDNA t-RFs digests (Block II).

 

 

 

Source PC1 PC2
df F Value P Value F Value P Value
Msp
Mosquito 1 1.46 0.246 37.95 <0.001*
Scirtid 1 3.09 0.099 0.86 0.368
Time 1 0.00 0.964 13.23 0.002*
Mosquito x Time 1 3.62 0.077 0.13 0.725
Scirtid*Time 1 1.01 0.331 0.07 0.798
Mosquito*Scirtid 1 1.76 0.204 2.27 0.153
Error 15
Hha
Mosquito 1 1.80 0.199 0.84 0.373
Scirtid 1 0.11 0.739 0.36 0.556
Time 1 0.10 0.758 6.37 0.023
Mosquito*Time 1 0.24 0.629 1 1.45 0.004*
Scirtid*Time l 0.00 0.956 0.13 0.723
Mosquito*Scirtid 1 1.72 0.208 0.2 0.663
Error 16

 

*Indicates p-values that are significant following Bonferroni adjustment

181

Appendix 2, Table 7. Summary of ANOVA results for PC1 and PC2 scores from leaf

surface 188 rDNA t-RFs digests (Block I).

—
Source PC1 PC2

df F Value P Value F Value P Value

 

 

Msp
Mosquito 1 0.08 0.785 0.80 0.390
Scirtid 1 0.51 0.492 0.83 0.381
Time 1 0.02 0.879 1.98 0.187
Mosquito x Time 1 0.16 0.698 0.19 0.673
Scirtid*Time 1 0.73 0.412 0.00 0.994
Mosquito*Scirtid 1 0.44 0.521 0.00 0.955
Error 1 1

Hha
Mosquito 1 0.77 0.400 0.34 0.574
Scirtid 1 0.29 0.601 0.17 0.687
Time 1 10.1 0.340 0.57 0.466
Mosquito*Time 1 1.21 0.297 0.19 0.675
Scirtid*Time l 0.08 0.782 0.20 0.665
Mosquito*Scirtid 1 2.13 0.175 0.06 0.809
Error 10

 

*Indicates p-values that are significant following Bonferroni adjustment to

182

Appendix 2, Table 8. Summary of ANOVA results for PC1 and PC2 scores from leaf

surface 188 rDNA t-RFs digests (Block II).

_
Source PC1 PC2

df F Value P Value F Value P Value

 

 

Msp
Mosquito l 0.05 0.829 6.19 0.024
Scirtid 1 0.23 0.639 1.20 0.289
Time 1 1.02 0.329 3.21 0.092
Mosquito x Time 1 0.16 0.692 1.18 0.2931
Scirtid*Time l 0.01 0.917 1.67 0.214
Mosquito*Scirtid 1 0.12 0.732 0.70 0.414
Error 16

Hha
Mosquito 1 1.58 0.226 4.19 0.058
Scirtid 1 0.35 0.560 0.01 0.922
Time 1 24.29 <0.001* 8.65 0.010*
Mosquito*Time 1 0.05 0.829 3.38 0.085
Scirtid*Time 1 0.68 0.421 1.29 0.272
Mosquito*Scirtid 1 1.40 0.254 0.02 0.867
Error 16

 

*Indicates p-values that are significant following Bonferroni

183

LITERATURE CITED

184

Literature Cited

Abdo, Z., U.M.E. Schfiette, S.J. Bent, C]. Williams, L.J. Fomey, and P. Joyce. 2006.
Statistical methods for characterizing diversity of microbial communities by analysis

of terminal restriction fragment length polymorphisms of 16S rRNA genes. Environ.
Microbiol. 8: 929—938.

Andreadis, T. G., J. F. Anderson, L. E. Munstermann, R. J. Wolfe, and D. A. Florin.
2001. Discovery, distribution, and abundance of the newly introduced mosquito

Ochlerotatusjaponicus (Diptera: Culicidae) inConnecticut, USA. J. Med. Entomol.
38: 774-779.

Bachmann, R.W. and Canfield, D. E. Jr. 1996. Use of an alternative method for

monitoring total nitrogen concentrations in Florida lakes. Hydrobiologia 323, 1-8.

Barlacher, F. 1985. The role of fungi in the nutrition of stream invertebrates. Botanical J.
Linnaen Society 91: 83-84.

Barrera, R. 1996. Competition and resistance to starvation in larvae of container-
inhabiting Aedes mosquitoes. Ecol. Entomol. 21: 117-127.

Bedhomme, S., P. Agnew, C. Sidobre and Y. Michalakis. 2003. Sex-specific reaction

norms to intraspecific larval competition in the mosquito Aedes aegypti. J. Evol.
Biol. 16: 721 -730.

Bomeman, J ., and Hartin, R. 2000. PCR primers that amplify fungal rRNA genes from
environmental samples. Appl. Environ. Microbiol. 66: 4356-4360.

Bradshaw, W.E., and Holzapfel, CM. 1992. Resource limitation, habitat segregation, and

species interactions of British tree-hole mosquitoes in nature. Oecologia. 90: 227-
237.

Braks, M. A. H., Honorio, N.A., Lourenco-de-Oliveira, R., Juliano S. A., and Lounibos
L. P. 2003. Convergent habitat segregation of Aedes aegypti and Aedes albopictus

(Diptera: Culicidae) in southeastern Brazil and Florida. J. Med. Entomol. 40:785-
794.

Braks, M. A. H., Honorio, N. A., Lounibos, L. P., Lourenco-de-Oliveira, R., and Juliano
S. A.. 2004. Interspecific competition between two invasive species of container

mosquitoes, Aedes aegypti and Aedes albopictus (Diptera: Culicidae), in Brazil.
Ann. Entomol. Soc. Am. 97: 130-39.

Calisher, C. H., 1983 Taxonomy, classification, and geographic distribution of California

serogroup Bunyaviruses, pp. 1—16 in California Serogroup Viruses, edited by C.
H. Calisher and W. H. Thompson. Alan R. Liss, New York.

185

Carpenter, S.R. 1982a. Comparisons of equations for decay of leaf litter in tree-hole
ecosystems. Oikos 39, 17-22.

Carpenter, S.R. 1982b. Stemflow chemistry effects on population dynamics of
detritivorous mosquitos in tree-hole ecosystems. Oecologia 53, 1-6.

Carpenter, SR. 1983. Resource limitation of larval treehole mosquitoes subsisting on
beech detritus. Ecology 64, 219-223. Carpenter, SR, and Kitchell, J. F. 1987. The
temporal scale of variance in lake productivity. Am. Nat. 129:417-433.

Carpenter, SR, and Kitchell, J. F. 1988. Consumer control of lake productivity.
BioScience. 38:764-769.

Centers for Disease Control and Prevention, 2007 http://wmwcdcgnv.

[CDC] Centers for Disease Control. Outbreak of West Nile-like viral encephalitis - New
York, 1999a. MMWR 1999, 48:845-849.

[CDC] Centers for Disease Control. Outbreak of West Nile-like viral encephalitis - New
York, 1999b. MMWR 1999, 48: 890-892.

Clements, A. N. 1992. “Development, nutrition, and reproduction.” The Biology of
Mosquitoes. Vol. 1. London: Chapman and Hall.

Craig, G. B., Jr. 1983. Biology of Aedes triseriatus: some factors affecting control. In:
California Serogroup Viruses. (ed. by C. H. Calisher and W. H. Thompson), pp.
329-341. Alan R. Liss, New York, New York.

Crumpton, W.G., Isenhart, TM. and Mitcell, PD. 1992. Nitrate and organic N analyses
with second-derivative spectroscopy. Limnology and Oceanography 37: 907-913.

Cummins, K.W. and Klug, M.J. 1979. Feeding ecology of stream invertebrates. Annual
Review of Ecology and Systematics 10, 147-172.

Dadd, RH. 1973. Insect nutrition - current developments and metabolic implications.
Ann. Rev. Entomol. 18: 381-420.

Daugherty, M. P., B. W. Alto and S. A. Juliano. 2000. Invertebrate carcasses as a
resource for competing Aedes albopictus and Aedes aegypti (Diptera: Culicidae).
J. Med. Entomol. 37: 364-372.

Daugherty, M. P., and Juliano, SA. 2001. Factors affecting the abundance of scirtid
beetles in container habitats. J. N. Am. Benthol. Soc. 20: 109-117.

186

Daugherty, M. P., and J uliano, SA. 2002. Testing for context-dependence in a processing
chain interaction among detritus-feeding aquatic insects. Ecol. Entomol. 27: 541 -
553.

Daugherty, M. P., and Juliano, SA. 2003. Leaf scraping beetle feces are a food resource
for tree hole mosquito larvae. Am. Midl. Nat. 150: 180-184.

Dixon, K.R., and Chapman, J.A. 1980. Harmonic mean measure of animal activity areas.
Ecol. 61: 1040-1044.

Dori go, U., L. Volatier and J .F . Hmnbert. 2005. Molecular approaches to the assessment
of biodiversity in aquatic microbial communities Water Res. 39: 2207—2218.

Edgerly, J. S., M. Mcfarland, P. Morgan and T. Livdahl. 1998. A seasonal shift in egg-
laying behaviour in response to cues of future competition in a treehole mosquito.
J. Anim. Ecol. 67: 805-818.

Fish, D. and Carpenter, SR. 1982. Leaf litter and larval mosquito dynamics in tree-hole
ecosystems. Ecology 63, 283-288.

Fish, D. 1983. Phytotelmata: Flora and Fauna. In: Phytotelmata: Terrestrial Plants as
Hosts for Aquatic Insect Communities (ed. by J. H. Frank and L. P. Lounibos),
pp. 1—27. Plexus, Medford, N.J.,

Gessner, M.W., and Schwoerbel, J. 1989. Leaching kinetics of fresh leaf-litter with
implications for the current concept of leaf-processing in streams. Archiv fuer
Hydrobiologie 115: 81-90. Gessner M.O., and Chauvet, E. 1993. Ergosterol-to-
biomass conversion factors for aquatic hyphomycetes. Appl. Environ. Microbiol.
59:502—507.

Graca, MA. 2001. The role of invertebrates on leaf litter decomposition in streams - a
review. Int. Rev. Hydrobiol. 86:383—393.

Grattan, R. M., and Suberkropp, K. 2001. Effects of nutrient enrichment on
yellow poplar leaf decomposition and fungal activity in streams. J. N. Am.
Benthol. Soc. 20:33—43.

Gubler, D. J ., Rosen, L. 1976. Variation among Geographic Strains of Aedes Albopictus
in Suceptibility to Infection with Dengue Viruses. Am. J. Trop. Med. Hyg. 25:
318-325.

Hard, J .J ., Bradshaw, W.E., and Malarkey, DJ. 1989 Resource— and density-dependent
development in tree-hole mosquitoes. Oikos 54: 137-144.

187

 

Hardy, J. L., E. J. Houk, L. D. Kramer, and W. C. Reeves. 1983. Intrinsic factors
affecting vector competence of mosquitoes for arboviruses. Annu. Rev. Entomol.
28:229-262.

Heard, SB. 1994. Pitcher-Plant Midges and Mosquitoes: A Processing Chain
Commensalism. Ecology. 75: 1647-1660.

Hobbs, J. H., E. A. Hughes, and B. H. Eichold 11. 1991. Replacement of Aedes aegypti
by Aedes albopictus in Mobile Alabama. J. Am. Mos. Contr. Assoc. 7 :488-489.

Homby, J. A,, D. E. Moore. and T. W. Miller, Jr. 1994. Aedes albopictus distribution,
abundance, and colonization in Lee County, Florida, and its effect on Aedes
aegypti. J. Am. Mos. Contr. Assoc. 10:397- 402.

J uliano, S. A. 1998. Species introduction and replacement among mosquitoes:
Interspecific resource competition or apparent competition? Ecology. 79:255-268.

J uliano, S.A., Lounibos, LP, and O 'Meara, OF. 2004. A field test for competitive
effects of Aedes albopictus on A. aegypti in South Florida: differences between
sites of coexistence and exclusion? Oecologia. 139: 583—593.

J uliano, S.A., and Lounibos, L. P. 2005.Ecology of invasive mosquitoes: effects on
resident species and on human health. Ecol. Lett. 8:558—5 74.

J firgens, K. and Matz, C. 2002. Predation as a shaping force for the phenotypic and
genotypic composition of planktonic bacteria. Antonie non Leeuwenhoek 81 :413-
434.

Kaufman, M.G., Walker, E.D., Smith, T.W., Merritt, R.W., and Klug, M.J. 1999. Effects
of larval mosquitoes (Aedes triseriatus) and stemflow on microbial community
dynamics in container habitats? Appl. Environ. Microbiol. 65: 2661-2673.

Kaufman, M.G., Bland, S.N., Worthen, M.E., Walker, ED, and Klug, M.J. 2001.
Bacterial and fungal biomass responses to feeding by larval Aedes triseriatus
(Diptera: Culicidae). J. Med. Entomol. 38: 71 1-719.

Kaufman, M.G., Goodfriend, W., Kohler-Garrigan, A., Walker, ED, and Klug, M.J.
2002. Soluble nutrient effects on microbial communities and mosquito production
in Ochlerotatus triseriatus habitats. Aquatic Microb. Ecol. 29, 73-88.

Kaufman, MC. and Walker, ED. 2006. Indirect effects of soluble nitrogen on growth of

Ochlerotatus triseriatus larvae in container habitats. J. Med. Entomol. 43: 677-
688.

188

Kaufman, M. G. and E. D. Walker. 2006. Indirect effects of soluble nitrogen on growth
of Ochlerotatus triseriatus larvae in container habitats. J. Med. Entomol. 43: 677—
688.

Kaufinan, M. G., S. Chen and E. D. Walker. 2008. Leaf-associated bacterial and fungal
taxa shifts in response to larvae of the tree hole mosquito, Ochlerotatus
triseriatus. Microb. Ecol. In press.

Kirchman, D. L. 2001. Measuring bacterial biomass production and growth rates from
leucine incorporation in natural aquatic environments. In: Methods in
Microbiology (ed by J. Paul), p. 227— 237. Academic Press, New York, NY.

Kirchman, D. L. 2002. The ecology of Cytophaga-Flavobacteria in aquatic environments
FEMS Microbio. Ecol. 39: 91—100.

Kitching, R.L. 1971. An ecological study of water-filled tree-holes and their positions in
the woodland ecosystem. J. Animal Ecol. 40: 281-302.

Kitching, R.L. 2001. Food webs in phytotelrnata: "Bottom-up" and "top-down"
explanations for community structure. Ann. Rev. Entomol. 46: 729-760.

Kneitel, J .M., and Miller, TE. 2002. Resources and top-predator regulation in the pitcher
plant Sarracenia purpurea inquinline community. Ecology 83: 680-688.

Laird, M. 1988. The natural history of larval mosquito habitats. San Diego CA, Academic
Press.

Lanciotti R. S, Roehrig J. T, Deubel V, Smith J, Parker M, Steele K, Crise B, Volpe K. E,
Crabtree M. B, Scherret J. H. 1999. Origin of the West Nile virus responsible for
an outbreak of encephalitis in the northeastern United States. Science 286:2333-
2337.

Leonard, P.M. and Juliano, SA. 1995. Effect of leaf-litter and density on fitness and
population performance of the hole mosquitoe Aedes triseriatus. Ecol. Entomol.
20: 125-136.

Liu W.T., Marsh T.L., Cheng H., and Fomey L.J. 1997. Characterization of microbial
diversity by determining terminal restriction fragment length polymorphisms of
genes encoding 16s rRNA. Appl. Environ. Microbiol. 63:4516-4522.

Livdahl, T. 1982. Competition within and between hatching cohorts of a treehole
mosquito. Ecology. 63: 175 1-1760.

Livdahl,T. P. and M. S. Willey. 1991. Prospects for an invasion: competition between
Aedes albopictus and native Aedes triseriatus. Science. 253: 189—191.

189

Lounibos, L. P., G. F. O’Meara, R. L. Escher, N. Nishimural, M. Cutwa, T. Nelson, R.E.
Campos and SA. J uliano. 2001. Testing predictions of displacement of native
Aedes by the invasive Asian tiger mosquito Aedes albopictus in Florida, USA.
Biol. Invasions. 3: 151—166.

Macia, A. and Brandshaw, W. E. 2000. Seasonal availability of resources and habitat
degradation for the western tree-hole mosquito, Aedes sierrensis. Oecologia. 125:
55-65.

Maknojia, S. 2006. Bacterial diversity and nutritional significance of the surface
microlayer in Anopheles gambiae (Diptera: Culicidae) larval habitats.
Dissertation.

Marchesi, J. R., T. Sato, A. J. Weightman, T. A. Martin, J. C. Fry, S. J. Hiom and W. G.
Wade. 1998. Design and evaluation of useful bacterium-specific PCR primers that
amplify genes coding for bacterial 16S rDNA. Appl. Environ. Microbiol. 64: 795-
99.

Marsh, T. L. 1999. Terminal restriction fragment length polymorphism (T-RFLP): an
emerging method for characterizing diversity among homologous populations of
amplification products. Cur. Opin. Microbiol. 2: 323-23.

McHugh, CR 1993. Distributional records for Aedes mosquitoes from the US. Air
Force ovitrapping program 1992. J. Am. Mosq. Cont. Assoc. 9:352—355.

Mekuria, Y., and M. G. Hyatt. 1995. Aedes albopictus in South Carolina. J. Am. Mosq.
Cont. Assoc. 11:468-470.

Menzel, D. W. and Corwin, N. 1965. The measurement of total phosphorus in seawater
based on the liberation of organically bound fractions by persulfate oxidation.
Limnol. Oceanogr. 10: 280-282.

Mercer, D. R. and J. R. Anderson. 1994. Tannins in treehole habitats and their effects on
Aedes sierrensis (Diptera: Culicidae) production and parasitism by Larnbomell
clarki (Ciliophora: Tetrahymenidae). J. Med. Entomol 31 : 159-167.

Merritt, R. W., R. H. Dadd, and ED. Walker. 1992. Feeding behavior, natural food, and
nutritional relationships of larval mosquitoes. Ann. Rev. Entomol. 37: 349-376.

Moorhead, D. L. and R.L. Sinsabaugh. 2006. A theoretical model of litter decay and
microbial interaction. Ecol. Monogr. 76: 1 5 1-174.

Murphy, J. and Riley, J .P. 1962. A modified single solution method for the detenninatin
of phosphate in natural waters. Analytica Chimica Acta 27: 31-36.

190

Newell, S.Y., and Barlocher, F. 1993. Removal of fungal and total organic matter from
decaying cordgrass leaves by shredder snails. J. Exp. Mar. Biol. Ecol. 171: 39-49.

Novak M. G., L. G. Higley, C. A. Christianssen, W. A. Rowley. 1993. Evaluating larval
competition between Aedes albopictus and A. triseriatus (Diptera: Culicidae)
through replacement series experiments. Environ. Entomol. 22: 311-18.

O'Meara. G. F .. A. D. Gettman, L. F. Evans. Jr., and G. A. Curtis. 1993. The spread of
Aedes albopictus in Florida. Am. Entomol. 39: 163-172.

O’Meara, G.F., Evans, L.F. Jr., Gettman, A.D., and Cuda, J .P. 1995 Spread of Aedes
albopictus and decline of Ae. aegypti (Diptera: Culicidae) in Florida. J. Med.
Entomol. 32: 554-562.

Paradise, C.J. 1999. Interactive effects of resources and a processing chain interaction in
treehole habitats. Oikos. 85: 529-535.

Paradise, C.J. 2000. Effects of pH and resources on a processing chain interaction in
simulated treeholes. J. Anim. Ecol. 69: 651-658.

Paradise, C.J. 2004. Relationship of water and leaf litter variability to insects inhabiting
tree holes. J. N. Am. Benthol. Soc. 23: 793-805.

Paradise, C.J., and Dunson, WA. 1997. Insect species interactions and resource effects
in treeholes: are helodid beetles bottom-up facilitators of midge populations?
Oecologia. 109: 303-312.

Pemthaler, J ., T. Posch, K. S'imek, J. Vrba, R. Amann, and R. Psenner. 1997.
Contrasting bacterial strategies to coexist with a flagellate predator in an
experimental microbial assemblage. App. Environ. Microbiol. 63: 596-601.

Peyton E. L, Campbell S. R, Candelletti T. M, Romanowski M, and Crans W. J. Aedes
(Finlaya) japonicus japonicus (Theobald), a new introduction into the United
States. J. Am. Mosq. Control Assoc. 1999, 15:238-241.

Porter, KG. 1996. Integrating the microbial loop and the classic food chain to a realistic
planktonic food web. In Food Webs, Inegration of Patterns and Dynamics (ed. by
G. A. Polis and K. O. Winemille), pp. 51-559. New York: Chapman and Hall.

Porter, K.G. and F eig, Y.S. 1980. The use of DAPI for identifying and counting aquatic
microflora. Limnol. Oceanogr. 25: 943-948.

Porter, K. G., and Y. S. F eig. 1980. The use of DAPI for identifying and counting aquatic
microflora. Limnol. Oceanogr. 25:943—948.

Rai, K. S. 1991. Aedes albopictus in the Americas. Annu. Rev. Entomol. 36:459-84.

191

R Development Core Team. 2004. R: A language and environment for statistical
computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-
900051-07—0, URL http://wwwR-proiectorg.

Reid R. Gerhardt, R.R., Gottfiied, K.L. Apperson, C.S., Davis, B.S. Davis, Erwin, P.C.,
Smith, A. B., Panella, N.A., Powell, E.E., and Nasci, R.S. 2002.First isolation of

La Crosse Virus from naturally infected Aedes albopictus. Emerg Infec Dis.
72807-811.

Riemann, B. and Azam, F. 1992. Measurements of bacterial protein synthesis in aquatic
environments by means of leucine incorporation. Marine Microbial Food Webs 6:
91-105.

Reinert J .F . 2000. New classification for the composite genus Aedes (Diptera: Culicidae:
Aedini), elevation of subgenus Ochlerotatus to generic rank, reclassification of the
other subgenera, and notes on certain subgenera and species. J. Am. Mosq.
Control Assoc. 16: 175-188.

Reinert J .F., Harbach RE, Kitching U. 2004. Phylogeny and classification of Aedini
(Diptera: Culicidae), based on morphological characters of all life stages. Zool. J.
Linnean Soc. 142: 289-368.

Rice, W.R. 1989. Analyzing tables of statistical tests. Evolution 43: 223-225.

Roppo, M., J. Lilja, F. Maloney and W. Sames. 2004. First occurrence of Ochlerotatus
japonicus in the state of Washington. J. Am. Mosq. Control Assoc. 20: 83—84.

Rossi, L. 1985. Interaction of invertebrates and microfungi in freshwater ecosystems.
Oikos 44:175—84.

Samuels G.J., and Blackwell, M. 2001. Pyrenomycetes-fungi with perithecia. In:
McLaughlin DJ, McLaughlin EG, Lemke PA, eds. The Mycota VII Part A. Berlin
Heidelberg: Springer-Verlag. p. 221-255.

Sardelis M. R, Turell M. J, Dohm D. J, and O'Guinn M. L. Vector competence of
selected North American Culex and Coquillettidia mosquitoes for West Nile
virus. Emerg. Infect. Dis. 2001, 7: 101 8-1022.

Sardelis M. R, Turell M. J, O' Guinn M. L, Andre R. G, and Roberts D. R. Vector
competence of three North American strains of Aedes albopictus for West Nile
virus. J. Am. Mosq. Control Assoc. 2002, 18:284-289.

SAS Institute Inc. 2000. SAS/STAT User’s Guide, version 4th ed. SAS Institute, Cary,
NC.

192

Scott, J .J ., F. L. Carle, and W. J. Crans. 2001. Ochlerotatusjaponicus collected from
natural rockpools in New Jersey. J. Am. Mosq. Control Assoc. 17: 91—92.

S'imek, K., J. Vrba, J. Pemthaler, T. Posch, P. Hartman, J. Nedoma, and R. Psenner.
1997. Morphological and compositional shifts in an experimental bacterial
community influenced by protists with contrasting feeding modes. App. Environ.
Microbiol. 63:587—595.

Smith, DC. and Azam, F. 1992. A simple, economical method for measuring bacterial
protein synthesis rates in seawater using 3H-leucine. Marine Microbial Food
Webs 6: 107-114.

Sardelis M. R. and M. J. Turrell. 2001. Ochlerotatus j. japonicus in Frederick County,
Maryland: discovery, distribution, and vector competence for West Nile virus. J
Am. Mosq. Control Assoc. 17:137-41.

Sommaruga, R., and Psenner R. 1995. Permanent presence of grazing resistant bacteria in
a hypertrophic lake. App. Environ. Microbiol. 61 :3457— 3459.

Sota, T., M. Mogi and E. Hayamizu. 1994. Habitat stability and the larval mosquito
community in treeholes and other containers on a temperate island. Res. Pop.
Ecol. 36293-104.

Suberkropp, K., and Weyers, H. 1996. Application of firngal and bacterial production
methodologies to decomposing leaves in streams. Appl. Environ. Microbiol. 62:
1 610-161 5.

Talbot, G., E. Topp, M.E. Palin, and DJ. Massé. 2008. Evaluation of molecular methods
used for establishing the interactions and functions of microorganisms in
anaerobic bioreactors. Water Res. 42: 513-537.

Tall L., Cattaneo, A. Cloutier, L., Dray, S., and Legendre, P. 2006. Resource partitioning
in a grazer guild feeding on a multilayer diatom mat. J. N. Am. Benthol. Soc.
25:800—810.

Teng, H. and C. S. Apperson. 2000. Development and survival of immature Aedes
albopictus (Diptera: Culicidae) in the laboratory: effects of density, food, and
competition on response to temperature. J. Med. Entomol. 37: 40-52.

Tryzcinski, M.K., Walde, S.J., and Taylor, PD. 2005. Stability of pitcher-plant
microfaunal populations depends on food web structure. Oikos 110: 146-154.

Turell, M. J ., Dohm, D.J., Sardelis, M. R., O'guinn, M.L., Andreadis, T.G., Blow, J. A.

2005. An Update on the Potential of North American Mosquitoes (Diptera:
Culicidae) to Transmit West Nile Virus. J. Med. Entomol. 42: 57-62.

193

Velgi, M. I., and L. J. Albright. 1993. Improved sample preparation for enumeration of
aggregated aquatic substrate bacteria. In: Handbook of Methods in Aquatic
Microbiology (ed. by P. F Kemp, B. F. Sherr, E. B. Sherr, and J. J. Cole.), pp. 139
-142 Lewis, Boca Raton, FL.

Watts, D. M., C. D. Morris, R. E. Wright, G. R. DeFoliart and R. P. Hanson, 1972
Transmission of La Crosse virus (Califomia encephalitis group) by the mosquito
Aedes triseriatus. J. Med. Entomol. 9: 125—127.

Webster, J .R. and Benfield, ER 1986. Vascular plant breakdown in freshwater
ecosystems. Ann. Rev. Ecol. Syst. 17: 567-594.

Walker, E. D. and R. W. Merritt. 1988a. The significance of leaf detritus to mosquito
(Diptera: Culicidae) productivity fiom treeholes. Environ. Entomol. 17 : 199—206.

Walker, E. D., E. J. Olds, and R. W. Merritt. 1988b. Gut content analysis of mosquito
larvae (Diptera: Culicidae) using DAPI stain and epifluorescence microscopy. J.
Med. Entomol. 25:551—554.

Walker, E.D., Lawson, D.L., Menitt, R.W., Morgan, W.T., and Klug, M.J. 1991. Nutrient
dynamics, bacterial-populations, and mosquito productivity in tree hole
ecosystems and microcosms. Ecology 72, 1529-1546.

Walker, ED. and Merritt, R.W. 1991. Behavior of larval Aedes triseriatus (Diptera,
Culicidae). J. Med. Entomol. 28, 581-589.

Walker, E. D., M. G. Kaufman, M. P. Ayres, M. H. Riedel, and R. W. Merritt. 1997.
Effects of variation in quality of leaf detritus on growth of the eastern tree—hole
mosquito, Aedes triseritatus (Diptera: Culicidae). Can. J. Zool. 75: 706-718.

Webster, J. R. and Benfield, ER 1986. Vascular plant breakdown in freshwater
ecosystems. Annual Review of Ecology and Systematics 17 567-594.

Xu, Y., S. Chen, M. G. Kaufman, S. Maknojia, M. Bagdasarian and E. D. Walker. 2008.
Bacterial community structure in tree hole habitats of Ochlerotatus triseriatus:
influences of larval feeding. J. Am. Mosq. Contr. Assoc. In press.

Yee, D. A., B. Kesavaraju and S. A. Juliano. 2004. Interspecific differences in feeding
behavior and survival under food-limited conditions for larval Aedes albopictus
and Aedes aegypti (Diptera: Culicidae). Ann. Entomol. Soc. Amer. 97: 720-728.

Yee, DA. and J uliano, SA. 2006. Consequences of detritus type in an aquatic

microsystem: effects on water quality, micro-organisms and performance of the
dominant consumer. Freshwater Biol. 51, 448-459.

194

Yee, D. A., M. G. Kaufman and S. A. Juliano. 2007 The significance of ratios of detritus
types and micro-organism productivity to competitive interactions between
aquatic insect detritivores. J. Animal Ecol. 76: 1105-1115.

Woodring, J. L., S. Higgs, and B. J. Beaty. 1996. Natural cycles of vector-borne
pathogens, p. 51-72. In B. J. Beaty, and W. C. Marquardt (ed), The biology of
disease vectors. University Press of Colorado, Niwot, Colo.

Zhang, Ning and Sung, Gi-Ho. 2008. Sordariomycetes. Version 14 January 2008 (under

construction). http://tolweb.org/Sordariomycetes/29050/2008.01.14 in The Tree
of Life Web Project, httpzl/tolweborg/

195