Louisiana State University LSU Digital Commons
LSU Historical Dissertations and Theses Graduate School
2000 Host and Parasite Interactions Among Three Symbionts: Pocket Gophers, Chewing Lice, and Endosymbiotic Bacteria. David Lee Reed Louisiana State University and Agricultural & Mechanical College
Follow this and additional works at: https://digitalcommons.lsu.edu/gradschool_disstheses
Recommended Citation Reed, David Lee, "Host and Parasite Interactions Among Three Symbionts: Pocket Gophers, Chewing Lice, and Endosymbiotic Bacteria." (2000). LSU Historical Dissertations and Theses. 7384. https://digitalcommons.lsu.edu/gradschool_disstheses/7384
This Dissertation is brought to you for free and open access by the Graduate School at LSU Digital Commons. It has been accepted for inclusion in LSU Historical Dissertations and Theses by an authorized administrator of LSU Digital Commons. For more information, please contact [email protected]. INFORMATION TO USERS
This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, som e thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer.
The quality of this reproduction is dependent upon the quality of the copy subm itted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction.
In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion.
Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand comer arid continuing from left to right in equal sections with small overlaps.
Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6” x 9n black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to ordesr.
Bell & Howell Information and Learning 300 North Zeeb Road, Ann Arbor, Ml 48106-1346 USA 800-521-0600
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. HOST AND PARASITE INTERACTIONS AMONG THREE SYMBIONTS: POCKET GOPHERS, CHEWING LICE, AND ENDOSYMBIOTIC BACTERIA
A Dissertation
Submitted to the Graduate Faculty of the Louisiana State University and Agricultural and Mechanical College in partial fulfillment of the requirement for the degree of Doctor of Philosophy
in
The Department of Biological Sciences
by David L. Reed B.S., University of North Carolina, Wilmington, 1991 M.S., Louisiana State University, 1994 December 2000
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. UMI Number: 9998703
UMI___ ®
UMI Microform 9998703 Copyright 2001 by Bell & Howell Information and Learning Company. All rights reserved. This microform edition is protected against unauthorized copying under Title 17, United States Code.
Bell & Howell Information and Learning Company 300 North Zeeb Road P.O. Box 1346 Ann Arbor, Ml 48106-1346
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ACKNOWLEDGEMENTS
I would like to thank my advisor, Mark S. Hafner, for the countless hours
that he invested in my education. I benefited enormously from his friendship,
guidance, and knowledge. My committee members, Mark S. Hafner, Fred H.
Sheldon, J. Michael Fitzsimons, J. V. Remsen, and William Font, encouraged me
at every step along the way, and for their help, I am grateful. Fred Rainey and
Naomi Ward-Rainey were very helpful in the areas of systematic and
evolutionary microbiology. I appreciate their expertise and willingness to share
their knowledge. Shannon Allen, Mark S. Hafner, and Melanie B. Smith
contributed significantly to the work presented in chapters one and two. I am
indebted to Dwanda Lewis, Myles Digby, Justin Lyman, and Rebecca Miller who
cloned and sequenced bacterial DNA for chapters three and four. I simply could
not have accomplished my research goals if not for these collaborators.
I would like to thank my parents for giving me the early education
necessary to excel in life. They taught me the importance of a well-rounded
education and I am forever grateful to them. I would like thank my wife, Jamie,
for tolerating a Ph.D. candidate who was more often at his office than at home.
Lastly, I'd like to thank one of the most important groups in my graduate training.
From my earliest mentors, Jim Demastes and Theresa Spradling, to my most
recent friends and colleagues, I thank the graduate students at the LSU Museum of
Natural Science. There is no more important source of knowledge and
inspiration in the Museum of Natural Science than its graduate students. A
special thank you to each of you. Lastly, I would like to thank Dr. Charles M.
Fugler for being a great mentor, friend, and colleague. He is deeply appreciated
and equally missed.
ii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. TABLE OF CONTENTS
ACKNOWLEDGEMENTS...... ii
LIST OF TABLES...... v
LIST OF FIGURES...... vi
ABSTRACT...... viii
INTRODUCTION...... 1
CHAPTER 1. MAMMALIAN HAIR DIAMETER AS A POSSIBLE MECHANISM FOR HOST SPECIALIZATION IN CHEWING LICE...... 4 Introduction...... 4 Materials and Methods...... 5 Results...... 10 D iscussion ...... 16
CHAPTER 2. SPATIAL PARTITIONING OF HOST HABITAT BY CHEWING LICE OF THE GENERA GEOMYDOECUS AND THOMOMYDOECUS (PHTHIRAPTERA: TRICHODECTIDAE)...... 23 Introduction...... 23 Materials and Methods...... 24 R esults...... 27 D iscussion...... 31
CHAPTER 3. BACTERIAL DIVERSITY IN CHEWING LOUSE HOSTS...... 36 Introduction...... 36 Materials and Methods...... 46 Results...... 52 D iscussion...... 58
CHAPTER 4. BACTERIAL DIVERSITY IN CHEWING LOUSE EGGS...... 67 Introduction...... 67 Materials and Methods...... 69 Results and Discussion...... 71
CHAPTER 5. STUDIES OF COPHYLOGENY: CHEWING LICE AND ENDOSYMBIOTTC BACTERIA...... 76 Introduction...... 76 Materials and Methods...... 78 Results...... 81 D iscussion...... 97
SUMMARY AND CONCLUSIONS...... 105
LITERATURE CITED...... 107
iii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. APPENDIX: SPECIMENS EXAMINED...... 117
VITA...... 119
iv
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LIST OF TABLES
Page
1.1 Model II regression analyses (major axis method) of the relationship between hair diameter (Diam) and body mass (Mass) ...... 11
2.1 Mean guard hair diameter (pm) for each of the ten body regions in the three pocket gopher specimens examined...... 29
2.2 Total surface area (cm2) of each pelage region (Fig. 2.1) and number of observed and expected lice per region...... 30
3.1 Bacteria associated with chewing lice identified to the level of genus by means of BLAST searches of GenBank ...... 53
3.2 Endosymbiotic bacteria identified from nine taxa of chewing lice by means of environmental extraction, PCR, sequencing, and searching GenBank databases via BLAST nucleotide recognition search protocol...... 56
4.1 Types of bacteria found in DNA extracts of chewing louse eggs. Bacteria were identified by BLAST searches of the Genbank DNA sequence database ...... 72
5.1 Uncorrected pairwise distances (P-distances) for 27 taxa of the gamma subclass of the Proteobacteria ...... 85
5.2 Uncorrected P-distances for 42 taxa of Staphylococcus...... 91
v
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LIST OF FIGURES
Page
1.1 Electron micrograph (left) of a chewing louse (Geomydoecus aurei) attached to the hair shaft of a pocket gopher (Thomomys bottae)...... 6
1.2 Major axis regressions performed on ln(mass) and In(diameter) for 18 species of mammals...... 12
1.3 Major axis regression through the origin of independent contrasts of ln(mass) and In(diameter) for members of the family Geomyidae ...... 13
1.4 Major axis regression through the origin of independent contrasts of ln(mass) and In(diameter) for pocket gophers in the genus Thomomys...... 14
1.5 Regression (major axis method) performed on In(mass) and In(diameter) for eight individuals of Thomomys bottae...... 15
1.6 Linear regression (general linear model) performed on chewing louse groove width and pocket gopher hair diameter...... 17
2.1 External surface of a pocket gopher showing the distribution of Geomydoecus and Thomomydoecus lice...... 26
2.2 Comparative density of lice in the 10 regions of pocket gopher pelage...... 32
3.1 Regression of the number of types of bacteria found as a function of the total number of clones sequenced for each of the seven gopher-Iouse host pairs listed in Table 3.2...... 60
3.2 Rarefaction curve showing the number of clones identified from Geomydoecus expansus compared to the total number of clones surveyed ...... 62
3.3 Rarefaction curve showing the number of clones identified from Geomydoecus geomydis compared to the total number of clones surveyed ...... 63
5.1 Maximum likelihood phylogeny for gamma Proteobacteria ...... 84
vi
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5.2 Fifty percent majority rule consensus tree of 1000 bootstrap replicates using the parsimony optimality criterion for species of gamma Proteobacteria...... 88
5.3 Maximum likelihood phylogeny for Staphylococcus species...... 90
5.4 Fifty percent majority rule consensus tree of 1000 bootstrap replicates using the parsimony optimality criterion Staphylococcus species...... 96
vii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ABSTRACT
This study investigates the patterns of cophylogeny documented
between pocket gophers (Rodentia: Geomyidae) and their ectoparasitic
chewing lice (Phthiraptera: Trichodectidae). The first two chapters
investigate the mechanisms that reinforce cospeciation in this system.
Chapter one examines whether mammalian hair diameter is a resource
tracked by chewing lice, and thus reinforces the patterns of cospeciation. I
found that hair diameter of pocket gophers is correlated with gopher body
size and the body size of the ectoparasitic chewing lice. Hair diameter also is
correlated with size of the rostral groove, which chewing lice use to grasp
hair shafts of pocket gophers. In chapter two I examine whether hair
diameter is the resource by which coexisting chewing louse species partition
available habitat. Although the pattern of spatial partitioning was clearly
evident, hair diameter does not seem to be the mechanism by which lice
partition their habitat. Alternative resources could include temperature,
humidity, or the distribution of sebaceous glands throughout the gopher
pelage. Chapters three, four, and five investigate the bacteria associated with
the pocket gopher-chewing louse system. I used the culture-independent
method of amplifying DNA sequences to determine the identification of
bacteria extracted from samples of chewing lice. This method yielded
approximately 35 distinct lineages of bacteria associated with this system. I
also amplified DNA sequences of bacteria from the eggs of chewing lice to
v iii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. determine whether any bacteria were transferred vertically in insects
through the eggs. Several species of bacteria were repeatedly seen in these
amplifications, which could be explained by the incorporation of bacteria in
the insect eggs. Further study would be required to provide conclusive
evidence of direct vertical transmission of bacteria in chewing louse eggs.
The final chapter of this dissertation looks in detail at two groups of bacteria
associated with chewing lice (gamma-Proteobacteria and Staphylococcus
species). Complete sequencing of thel6S rRNA gene demonstrated that
multiple species of both groups are associated with the pocket gopher-
chewing louse system. Studies of cophylogeny between the species of
Staphylococcus and their hosts were inconclusive and require more
intensive taxon sampling.
ix
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. INTRODUCTION
So, naturalists observe, a flea
Has smaller fleas that on him prey;
And these have smaller still to bite 'em;
And so proceed ad infinitum.
Jonathan Swift (1667-1745) in Poetry, a Rhapsody.
Since 1984, Mark Hafner and his colleagues have studied cophylogeny
(also termed cospeciation) in a group of rodents (pocket gophers) and their
insect ectoparasites (chewing lice; for a review of this work see Hafner,
Demastes, Spradling, and Reed in press). One portion of my research
explores the possible mechanisms that may have reinforced if not caused the
pattern of cospeciation documented in this host-parasite system. In chapter
one, I examine the relationship between mammalian hair diameter, body
mass, and chewing louse rostral groove diameter. I show a significant,
positive allometric relationship between mammalian hair diameter and
body mass. I also show a significant positive relationship between pocket
gopher hair diameter and the rostral groove dimensions of chewing lice.
Lice use this rostral groove to grasp hairs of their host. Coupled with
previous evidence of a strong allometric relationship between rostral groove
width and louse body size, these findings suggest that hair diameter of the
host is an important determinant of body size in chewing lice. When
1
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. viewed in light of the extreme host specificity of chewing lice (Reed and
Hafner 1997), one can postulate that the fit between gopher hair diameter
and louse rostral groove diameter may reinforce patterns of cospeciation.
Certain taxa of chewing lice of the genera Geomydoecus an d
Thomomydoecus coexist on pocket gophers of the genus Thomomys. In
chapter two, I investigate the spatial distribution of coexisting louse species
on their hosts and explore possible mechanisms of resource partitioning by
chewing lice. Chewing lice appear to partition available host resources
spatially, with Geomydoecus occurring primarily on the lateral and dorsal
regions of the host, and Thomomydoecus occurring primarily on the lateral
and ventral regions. Although spatial partitioning of the host habitat is
evident, it does not appear to be explained by hair diameter. Ecological and
behavioral interactions between these two genera of chewing lice may have
important effects on gopher-louse cospeciation.
My research extended the gopher-louse study of cophylogeny to a
third taxonomic group, the endosymbiotic bacteria of chewing lice. The
asociality and patchy distribution of pocket gophers, compounded by the
extreme isolation of chewing louse populations on individual hosts, suggest
that the bacterial endosymbionts of chewing lice may show patterns of
geographic variation that mirror those of their hosts. Many biologists
assume, either implicitly or explicitly, that microbial species are ubiquitous
(or nearly so) and that chance plays a major role in determining microbial
community structure at any specific locality (e.g. Ward et al. 1998). Recent
2
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 16S rRNA-based molecular surveys of microbial communities seem to
support the assumption of bacterial ubiquity, in that organisms with
similar—sometimes identical—16S rRNA sequences have been recovered
from geographically distant localities (Staley 1999).
My research has examined this assumption in the context of a well-
defined host-parasite system by testing the alternative hypothesis that
bacterial lineages show phylogenetic structuring that parallels that of their
hosts (cophylogeny). Chapter three is an examination of the diversity of
bacteria found within chewing lice that parasitize pocket gophers. Many of
these bacteria likely are short-term transients in the gopher-louse system,
perhaps acquired by the lice through diet. In contrast, other bacteria may be
obligate endosymbionts that perform functions necessary for louse survival.
Many endosymbiotic bacteria are transferred from female insects to their
progeny via the insect ovum. Chapter four describes the search for bacteria
in the eggs of chewing lice. Transovarial inoculation of insects is a complex
mechanism that likely evolved slowly. Therefore, bacterial species found in
the eggs of chewing lice likely are long-term associates of the lice and may
show patterns of cophylogeny with their louse hosts. The fifth, and final
chapter, is a study of cophylogeny between chewing lice and a selected group
of endosymbiotic bacteria of the genera Staphylococcus and Acinetobacter.
The results of this study are discussed in the context of those from the
previous four chapters in a closing section titled "Summary and
Conclusions".
3
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 1
MAMMALIAN HAIR DIAMETER AS A POSSIBLE MECHANISM FOR HOST SPECIALIZATION IN CHEWING LICE
INTRODUCTION
Body size of host and parasite often are correlated positively. Harvey
and Keymer (1991) and Morand et al. (2000) demonstrated this trend for
chewing lice (Phthiraptera: Trichodectidae) and their pocket gopher hosts
(Rodentia: Geomyidae). Specifically, Harvey and Keymer (1991) used the
comparative method to demonstrate that increased body size in pocket
gophers is associated invariably with increased size of their ectoparasites.
They suggested that lice grow larger on larger hosts because those hosts
presumably live longer and allow their lice more time to grow. Although
intriguing, this explanation seems unlikely, given that generation time of
chewing lice (about 40 days; Rust 1974) is almost an order of magnitude less
than generation time of even the shortest-lived species of pocket gopher
(about one year; Nowak 1999).
Other potential explanations exist for the body-size correlation
documented by Harvey and Keymer (1991). For example, ability of the host
to detect and destroy ectoparasites may scale with host body size and thereby
place an upper limit on body size of the parasite. Evolutionary changes in
the body mass of the host also might alter habitat of chewing lice in terms of
temperature, humidity, hair length and diameter, and other habitat
4
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. parameters. Because lice are extremely host specific (Price and Emerson
1972; Reed and Hafner 1997) and inextricably tied to their host for survival
(Kellogg 1913; Marshall 1981), it seems likely that lice would show finely
tuned adaptations for life on their host. Characteristics of the hair,
particularly hair diameter, should be important components of the
environment of the louse (Fig. 1.1) because of the pivotal role that hair plays
in louse feeding, locomotion, ovipositing, and survival (Murray 1957a;
1957b).
Mammal pelage consists of two basic types of hairs: guard hairs,
which are relatively long and thick, and wool hairs (or underfur), which are
shorter, thinner hairs (Mayer 1952). My observations of chewing lice reveal
that lice spend most of their time moving among guard hairs and are
seldom seen attached to wool hairs. Accordingly, I focused this
investigation on mammalian guard hairs. I examined the relationship
between the diameter of guard hair and body mass at several taxonomic
levels in mammals (interordinal, intrafamilial, intrageneric, and
intraspecific) to determine the generality of the hair size-body size
relationship in mammals. I also investigated whether hair diameter of the
host was correlated with host and parasite body size in geomyid rodents.
MATERIALS AND METHODS
A single adult individual from each of 18 species of non-geomyid
mammals (representing 17 families in 9 orders; Appendix I) was
5
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Fig. 1.1. Electron micrograph (left) of a chewing louse {Geomydoecus aurei) attached to the hair shaft of a pocket gopher (Thomomys bottae). Magnified view (right) of rostral groove and hair shaft.
6
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. examined to assess variation in hair diameter across different orders of
mammals. Taxa were selected randomly from the Louisiana State
University Museum of Natural Science Collection of Mammals, except that
species with conspicuous spines or quills were avoided and range of body
size was maximized as much as possible (Appendix I). Because of difficulty
in sampling so many widely divergent taxa, some specimens were not
collected from the wild (Appendix I) and season of collection was not
standardized.
Guard hairs (n = 20 per individual) were removed from the nape region of
museum study skins. The nape region was selected to standardize the
sampling procedure and reduce the likelihood of damage to haiirs from
grooming. Hairswere mounted on microscope slides with Permount® and
secured with cover slips. Mathiak (1938) determined that the greatest
diameter of most mammalian guard hairs was found roughly one-half the
distance from the root to the tip. Accordingly, I measured hair diameter
about midway between the root and tip by using a light microscope fitted
with an ocular micrometer scale. Mass of each mammal in this study was
taken directly from the specimen tag or was estimated based on inform ation
provided by Nowak (1999).
Fourteen species in the family Geomyidae (Appendix I) were
examined to assess variation in hair diameter within a single family of
mammals. Guard hairs (n = 200 per specimen) were sampled from
throughout the gopher pelage and prepared for light microscopy. A single
7
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. individual from each of seven species in the genus Thom om ys (A ppendix I)
was examined to assess variation in hair diameter within a single genus of
pocket gophers. Guard hairs (n = 200 per specimen) were sampled from
throughout the pelage and prepared for microscopy. Likewise, eight
individuals from a single species, T hom om ys bottae, were examined to
assess variation in hair diameter within a single species of pocket gopher.
Guard hairs (n = 200 per specimen) were sampled from throughout the
gopher pelage and prepared for microscopy.
Width of the rostral groove was measured for adult lice collected
from the same individual gophers from which hair diameter was measured
and with about equal representation of male and female lice. Louse samples
were: Geomydoecus scleritus (n = 11) fro m Geomys pinetus; Geomydoecus
panamensis (n = 11) from Orthogeomys cavator; Geomydoecus setzeri (n = 8)
from O. underwoodi; Geomydoecus aurei (n = 10) from Thomomys bottae;
Thomomydoecus minor (n = 6) from T hom om ys bottae; and Geomydoecus
oregonus (n = 11) from Thomomys bulbivorus. Lice were cleared for light
microscopy by soaking in the following series of solutions (10-20
min/solution): 50% EtOH, 60% EtOH, 70% EtOH, 10% KOH, 80% EtOH, 90%
EtOH, 100% EtOH, and xylene. Lice were mounted on microscope slides,
secured with a coverslip, and allowed to dry for 24 h. Rostral groove width
was measured with a light microscope fitted with an ocular micrometer.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Because comparisons across species boundaries potentially are
confounded by phylogenetic relationships, independent contrasts
(Felsenstein 1985; Harvey and Pagel 1991) were used to remove phylogenetic
effects from these data. The computer program CAIC (Comparative
Analysis of Independent Contrasts; Purvis and Rambaut 1995) uses a
phylogenetic hypothesis to generate independent contrasts of data, which are
then analyzed statistically. Because most mammalian ordinal relationships
are unclear, comparisons among orders of mammals were not transformed
using CAIC. Such comparisons generally are considered independent
because the phylogenetic distance between terminal taxa was large. Hair
diameter and body mass measurements for pocket gophers were
transformed into independent contrasts by using a composite geomyid
phylogeny based on phylogenetic studies by Hafner et al. (1994), Smith (1998),
and Spradling (1997). No phylogenetic hypotheses were available for taxa
below the level of species; therefore, comparisons within Thomomys bottae
were not transformed into independent contrasts. Model II regression
analyses (major axis method) were performed with the SYSTAT statistical
analysis software package (SYSTAT, Inc. 1992). Model II regression is
appropriate when two variables lack a clear dependent-independent
relationship and both are measured with error (LaBarbera 1989; Martin and
Barbour 1989; Silva 1998). Regressions of independent contrasts were
constrained through the origin, as required by CAIC to retain n - 2 degrees of
freedom (Garland et al. 1992; Purvis and Rambaut 1995).
9
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. RESULTS
The regression analysis of hair diameter and body mass for 18 species
representing 9 orders of mammals (Fig. 1.2) revealed a positive relationship
(P < 0.05; Table 1.1). The allometric coefficient of this relationship (a = 0.13 ±
0.03) was low, which indicates that body mass increased more rapidly than
hair diameter among mammals examined, which ranged in body size from
a 3.5-g bat ( Pipistrellns) to a 600-kg bear ( Ursus).
Regression analysis of independent contrasts of hair diameter and
body mass for 14 species of pocket gophers (Fig. 1.3) revealed a similar
allometric trend (a = 0.25 ± 0.05; P < 0.05; Table 1.1). Thus, when analyzed at
the family level (and controlling for phylogenetic relationships within the
family), larger species of pocket gophers tend to have thicker guard hairs.
Examination of the relationship between hair diameter and body
mass within a single genus of pocket gophers (Thom om ys; Fig. 1.4) show ed
the same trend evident in the analyses at higher taxonomic levels. In
individuals representing seven species ofThomomys , hair diameter showed
a significant allometric relationship with body mass (a = 0.20 ± 0.07; P < 0.05;
Table 1.1).
Regression analysis of hair diameter and body mass for eight
individuals of Thomomys bottae (Fig. 1.5) also showed a significant, positive
relationship (a =0.33 ± 0.04; P < 0.05; Table 1.1). Thus, larger individuals of
10
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. <0.05 Cl P 95% 0.24-0.43 <0.05 0.02-0.37 <0.05 0.13-0.36 <0.05 0.06-0.19 0.03 SE 0.13 0.25 0.05 ln(Mass) 0.33 0.04 * Regression model Slope ln(Diam) = B * ln(Mass) 0.20 0.07 ln(Diam) = A + B * ln(Mass) (8) ln(Diam) = A + B (7) Taxon (») Table 1.1. Model II regression analyses (major axis method) of the relationship between hair diameter (Diam) and Thomomys bottae Thomomys Geomyidae (14) ln(Diam) = B * ln(Mass) Non-geom yid Mammalia (18) listed in Appendix I). body mass (Mass) at four taxonomic levels in mammals (depicted graphically in Figs. 1.2 through 1.5; taxa examined are
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5
=4.s 4
S-, jU 3 "S 6 .5 ■5 t - i 1 2 4 6 8 1 0 1 2 1 4 Body mass (in In g) Fig. 1.2. Major axis regressions performed on ln(mass) and ln(diameter) for 18 species of mammals. Regression models are in Table 1.1, and taxa examined are listed in Appendix I. 12 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 0 .3 re ■ J G u - i — .. ° g 0.2 cni C302 r-C U i — ■ c c o cr- ° fll u 0.1 g a -o £ fljc »p^ is O.T5 0.1 0 0.1 0.2 0 .3 0 .4 0 .5 Independent contrast of body mass (in In g) Fig. 1.3. Major axis regression through the origin of independent contrasts of ln(mass) and ln(diameter) for members of the family Geomyidae. Regression models are in Table 1.1, and taxa examined are listed in Appendix I. 13 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. • • 0.08 - 04 - 0 TJ -0.04 - -0.08 0 0.1 0.2 0.3 0.4 0.5 Independent contrasts of body mass (in In g) Fig. 1.4. Major axis regression through the origin of independent contrasts of ln(mass) and In(diameter) for pocket gophers in the genus Thomomys. Regression models are in Table 1.1, and taxa examined are listed in A ppendix I. 14 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4 3.5 3 2.5 3 4 5 6 Body mass (in In g) Fig. 1.5. Regression (major axis method) performed on ln(mass) and In(diameter) for eight individuals of Thomomys bottae. Individuals designated by squares (collected in native desert-scrub habitat) and triangles (collected in alfalfa fields) are those analyzed by Patton and Brylski (1987) and Smith and Patton (1988). These individuals are discussed further in the text. Regression models are in Table 1.1, and taxa examined are listed in Appendix I. 15 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. T. bottae tend to have thicker guard hairs than smaller individuals of the same species. The regression analysis of louse groove width and gopher hair diameter (Fig. 1.6) revealed a positive relationship between those variables (P < 0.05). The regression coefficient (a = 1.09 ± 0.09) suggested a near isometric relationship between groove width and hair diameter, which indicates that the two structures varied proportionately. Importantly, the Y- intercept of the regression (Fig. 1.6) w as close to zero ((3 = -3.9 /xm), w h ic h suggests that width of the rostral groove of a chewing louse was very similar in actual dimensions to maximum width of the guard hairs of its host. DISCUSSION Hair Diameter and Body Size My analyses document a consistent negative allometric relationship between hair diameter and body mass in mammals, regardless of the taxonomic level examined. The low allometric coefficient of this relationship (ranging from a = 0.13 to 0.33; Table 1.1) indicates that larger mammals tend to have guard hairs that are larger in absolute diameter but proportionately smaller than guard hairs of smaller mammalian species. This consistent relationship between hair diameter and body size in mammals is reminiscent of the negative allometric relationship observed for many other mammalian features that scale with body size (e.g., brain size, longevity, metabolism) and suggests that in most species of mammals hair diameter may be constrained within certain boundaries by simple 16 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (30c 70 • v*-O 'e' £ 03U , 3 — e 03 E 3 C8 O I—I 20 20 40 503060 70 80 90 Hair Diameter of Pocket Gopher (^m) Fig. 1.6. Linear regression (general linear model) performed on chewing louse groove width and pocket gopher hair diameter for six species of lice collected from five species of pocket gophers Geomys ( pinetus, Orthogeomys cavator, O. underxvoodi, Thomomys bottae, and T. bulbivorus). Regression models are shown in Table 1.1, and taxa examined are listed in Appendix I. 17 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. growth laws. This is not to suggest, however, that hair lacks important functional significance in most mammalian species or that adaptation has not played an important role in evolution of specialized hairs, such as spines or quills (specifically excluded from this analysis). Although hair diam eter, per se, may be rigidly constrained in many (or most) species of mammals, other aspects of mammalian pelage, including detailed microstructure of the hair, hair length, shape, color, and density (number of hairs per follicle and follicle density) are less likely to scale with body size and, thus, may have evolutionary flexibility in a wide variety of adaptive contexts. The analysis of hair diameter and body-mass relationships within a single species of pocket gopher (T. bottae; Fig. 1.5) em phasizes the tig h t linkage between these two variables in this species. This relationship suggests that a change in mean body mass within a lineage of pocket gophers over time (e.g., in response to climate change or an increase or decrease in food resources) will be accompanied by a corresponding change in hair diameter. To test this prediction, I included in my analysis six individuals of T. bottae that also were examined by Patton and Brylski (1987) and Smith and Patton (1988). Those studies compared body-size relationships between gophers collected from native desert-scrub habitat between 1937 and 1971 (indicated by squares in Fig. 1.5) to those collected from the same geographic region during 1984 and 1985, many years after the native vegetation had been converted to irrigated alfalfa fields (indicated by triangles in Fig. 1.5). 18 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Smith and Patton (1988) showed that the pocket gophers collected in the alfalfa fields (presumably direct descendants of the populations sampled decades earlier in native habitat) were significantly larger than their ancestors in overall body mass and other body dimensions. This increase in body mass is evident in my analysis (Fig. 1.5), and it is clear that the increase in overall body size in these pocket gophers was accompanied by an increase in hair diameter. Hair Diameter and Parasite Size Larger pocket gophers host larger chewing lice (Harvey and Keymer 1991; M orand et al. 2000), and the above analyses corroborate the prelim inary evidence presented by Morand et al. (2000) showing that larger pocket gophers also have thicker guard hairs. To establish a meaningful connection between hair diameter and louse body size, it is important to focus on some aspect of the body of the louse that interacts directly with the hair of the gopher. For this, I chose the rostral groove of the chewing louse, which is used to grasp the hair of the gopher (Fig. 1.1). Given that the rostral groove is a rather rigid structure, I predicted a close fit between rostral groove width in chewing lice and hair diameter in pocket gophers under the assumption that a louse with a very narrow rostral groove would be unable to grasp a thick hair, whereas a louse with a very wide groove may slip from a narrow hair. Murray's (1957a, 1957b) studies of chewing lice ( Damalinia ) on sheep (Ovis) showed that lice kept in laboratory colonies in which thin glass fibers were used as artificial hairs were extremely sensitive to width of 19 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. the glass fibers. If fibers were too thick, lice were unable to grasp them between the gonopod and abdomen and could not lay eggs. Similarly, differences in hair shape (and perhaps diameter) between humans of African descent and those of European descent have been used to explain reduced susceptibility of Africans to head lice (Pediculits humanus) from Europe and reduced susceptibility of Europeans to head lice of the same species from Africa (United States Centers for Disease Control 1984). Presumably, lice in the two regions have evolved differential abilities to grasp flat hairs typical of humans of African descent versus round hairs typical of humans of European descent. My results corroborate preliminary findings of Morand et al. (2000), which show a close fit between hair diameter in pocket gophers and rostral groove width of chewing lice when analyzed at interspecific and intergeneric levels (Fig. 1.6). This finding, coupled with Morand et al.'s (2000) discovery of a relationship between body size and groove width in chewing lice, suggests that interspecific variation in body size of lice may be determined largely by interspecific variation in diameter of gopher hair. If small species of lice (with narrow rostral grooves) are unable to grasp thick hairs of large pocket gophers and if large lice tend to avoid hosts with thin hairs, this could help explain the high level of host-specificity observed in chewing lice at zones of contact between gopher species of very different body sizes (M. S. Hafner, pers. comm.). This potentially obligate relationship between groove size and hair diameter also may explain why certain species of lice are 20 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. unable to survive on certain species of hosts in laboratory transfer experiments. For example, experiments by Reed and Hafner (1997) have shown that chewing lice transferred between species of pocket gophers of similar body size often are able to establish successful breeding colonies on foreign (non-native) hosts. Lice that occur naturally on a large species of host occasionally are able to survive and reproduce on a smaller species of host, but the reverse does not seem to be true. This suggests that the rostral groove of lice from the smaller hosts may be too narrow to grasp thick hairs of the larger host species. A similar study of body-size relationships between bird lice and their hosts has shown that feather size is a crucial factor in determining success of lice experimentally introduced onto new host taxa (D. Clayton, pers. comm.). Preliminary analyses of individual, sexual, and ontogenetic variation in hair diameter in pocket gophers (chapter two) suggest that variation at these levels may be within the normal range of tolerance for louse rostral grooves. This pattern would explain why a young pocket gopher does not "outgrow" its parasites as the gopher increases in body size (and, thus, hair diameter) ontogenetically. The pattern also would explain why male and female pocket gophers, which often show marked sexual dimorphism in size (Patton and Brylski 1987), nevertheless host the same species of chewing louse. My analyses demonstrate a negative allometric relationship between hair diameter and body mass among mammals when analyzed at multiple 21 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. taxonomic levels. This relationship is evident in the family Geomyidae, even at the intraspecific level. These data also reveal a significant positive relationship between hair diameter of pocket gophers and rostral groove width of their chewing lice. In fact, I show a nearly exact fit between hair diameter and groove width for host-parasite pairs. Given that lice eventually die if removed from their host, it seems likely that groove width would be under strong selective pressure to conform to host hair diameter. If groove width in lice scales with overall body size, then selection for optimal groove width could indirectly constrain body size in lice. However, the causal mechanism driving the positive relationship between host and parasite body size (as originally documented by Harvey and Keymer 1991) remains untested. Further studies involving direct observation of chewing lice transferred to non-native hosts may elucidate the causal mechanism underlying the empirical observation that larger species of pocket gophers tend to host larger species of chewing lice. 22 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 2 SPATIAL PARTITIONING OF HOST HABITAT BY CHEWING LICE OF THE GENERA GEOMYDOECUS AND THOMOMYDOECUS (PHTHIRAPTERA: TRICHODECTIDAE) INTRODUCTION Ectoparasitic chewing lice, Geomydoecus and Thomomydoecus spp. (Phthiraptera: Trichodectidae), live their entire lives exclusively on pocket gophers of the rodent family Geomyidae (Marshall 1981; Hellenthal and Price 1984). Chewing lice are wingless, obligate parasites that can survive only a short time when removed from their host (Kellogg 1913; Marshall 1981). One species of louse often is confined to a single species of host (Emerson and Price 1981), which suggests a long-term, perhaps obligate, association between each host-parasite pair. In many instances, this long term association has resulted in parallel cladogenesis between the pocket gopher and chewing louse lineages. This pattern, termed "cophylogeny," is well documented for certain lineages of pocket gophers and their associated chewing lice (Hafner and Nadler 1988; Demastes and Hafner 1993; Hafner et al. 1994). Although most individual pocket gophers host populations of a single species of louse, three species within Thomomys host representatives of two genera of chewing lice (Geomydoecus and Thomomydoecus; Hellenthal and Price 1984), with species of both genera usually coexisting o n an individual host. The principle of competitive exclusion (Gause 1934), 23 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. suggests that stable coexistence of two species of lice on an individual host relies on partitioning some aspect of the resources provided by the host (Durden 1987). During my initial studies of chewing louse distribution on pocket gophers, I noticed that guard hair diameter seems to vary predictably with body region, and perhaps provides a mechanism for resource partitioning by chewing lice. Lice attach to gopher hairs by means of a head groove located on the rostrum (Fig. 1.1). Secure attachment is necessary for survival of the parasite because of the louse's absolute dependence on the host. In chapter one of this dissertation I show a significant positive relationship between hair diameter in several genera of pocket gophers and rostral groove width of their chewing lice, and I also demonstrate that the head groove of a louse is of the appropriate size to grip tightly onto the hair shaft of its natural host. In the present study, I investigate hair diameter as a potential mechanism for resource partitioning that may result in stable coexistence between two species of chewing lice on an individual pocket gopher. Specifically, I test the hypothesis that louse species of Geomydoecus and Thomomydoecus are able to coexist by partitioning the host's resources spatially based on hair diameter. MATERIALS AND METHODS Three specimens of Thomomys bottae connectens w ere trapped on 17 March 1997 in Albuquerque, Bernadillo County, New Mexico. Two males and one female were collected with traps designed by Baker and Williams 24 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (1972). Pocket gophers were collected on a single day from a single locality to limit variation in louse population density caused by weather patterns, reproductive condition of the host, or other seasonal or geographic factors. Each gopher was killed by placing it in an air-tight container saturated with chloroform. This process also immediately immobilized and killed the resident louse population of each gopher. The gopher was then incised medially along the abdomen, and the entire skin was removed and pinned to a piece of cardboard, with the fur side down. Unnecessary movement of the skin was avoided to reduce accidental displacement of the lice. The skin was frozen on a block of dry ice, then pressed between two pieces of cardboard, wrapped tightly in aluminum foil and frozen in an ultracold freezer (-75° C). W hile frozen, the gopher skin was cut into 10 regions (Fig. 2.1): anterior ventral, cheek, dorsal head, lateral nape, lateral, nape, posterior dorsal, posterior ventral, rump, and ventral head. Samples from the right and left sides of the body were pooled for each region, and each region was placed individually in a plastic bag to avoid loss of lice and contamination by lice from other regions. Each section of the gopher pelt was brushed vigorously, and lice were collected in a 1.5-ml cryotube. Adult lice were then identified, using a dissection microscope, as either Geomydoecus aurei or Thomomydoecus minor. Only adults were used in this analysis because visual identification of juvenile lice is problematic. Each pelage region was measured (in cm2) so that the number of lice per region could be 25 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. lice. Thomomydoecus Head Ventral and Ventral Ventral Anterior Posterior 'Literal Geomydoecus Nape Rump Head Dorsal Dorsal Posterior Check Nape Lateral Lateral Ventral Ventral Anterior Head Posterior Ventral Head Ventral Ventral Ventral Anterior Posterior Nape Lateral Lateral Check Rump Head Nape Dorsal Dorsal Posterior Ihcck Nape Lateral Lateral Geomydoecus Thomomydoecus Ventral Ventral Head Anterior Posterior Ventral Figure 2.1. External surface of a pocket gopher showing the distribution of Darkly shaded regions contained morecontained lice than fewer expected lice than in allexpected three pocketin at leastgopher 2 of specimensthe pocket examinedgopher specimens(Table 2.1). examined. Lightly shaded regions contained more lice than expected in 2 of the 3 pocket gopher specimens examined. Unshaded regions Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. standardized (lice/cm 2). The total surface area (the 10 regions combined) and overall louse density (lice/cm2) were calculated for each of the three gopher specimens. If one assumes a null model of even louse distribution, the expected number of Geomydoecus and Thomomydoecus for each region based on the size of the region and the mean density for that particular gopher was determined. Chi-square analyses were used to test whether the observed numbers of lice were significantly different from expected numbers based on an even distribution. Ten guard hairs were taken from each of the 10 regions from all three gophers (chewing lice normally grasp guard hairs rather than underfur; pers. obser.). The hairs were mounted on microscope slides, and the mid-point diameter of each hair was measured (in ym) with a light microscope fitted with an ocular micrometer. Analysis of variance (ANOVA) and a Duncan post-ANOVA test (SAS Institute 1994) were used to detect significant differences in mean hair diameter among the ten regions. Specimens are deposited in the New Mexico Museum of Natural History (NMMNH 2378, 2379, and 2380). RESULTS Mean hair diameter (pooled data for the three hosts) varied from 33.61 ± 1.75 ym in the rump region to 45.22 ± 1.53 ym in the lateral-nape region (Table 2.1). The Duncan post-ANOVA test identified five groups of regions within which mean hair diameter was not significantly different 27 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (Table 2.1). In general, pocket gopher hair was smaller in diameter on the dorsal surfaces and larger in diameter on lateral and ventral surfaces. The overall mean density of Geomydoecus aurei (pooled data for all three hosts; Table 2.2) was 0.34 lice/cm2 (range 0.28-0.50 lice/cm2), and overall mean density for Thomomydoecus minor was 0.94 lice/cm2 (range 0.28-1.38 lice/cm2). Chi-square analysis of louse distribution revealed that none of the populations of either species was distributed evenly over the gopher pelage (all chi-square values exceeded the critical value of 27.88, P<0.001, df = 9; Table 2.2). Individuals of Geomydoecus aurei were found in all 10 regions of the gopher pelage, although some regions contained very few individuals (see pooled data, Table 2.2). Likewise, at least one individual of Thomomydoecus minor occurred in all regions except the dorsal-head region. Geomydoecus aurei occurred in greater abundance than expected on dorsal and lateral surfaces of the hosts (Fig. 2.1a), whereas Thomomydoecus minor was found in greater abundance than expected on lateral and ventral surfaces (Fig. 2.1b). For Geomydoecus, the regions of high abundance (shaded regions in Fig. 2.1a) comprised only 34% of the total surface area of the gopher yet contained 78% of all Geomydoecus individuals (Table 2.2). In contrast, the regions of high abundance for Thomomydoecus (shaded regions in Fig. 2.1b), which also comprised 34% of the total surface area of the gopher, contained only 55% of all Thomomydoecus individuals. 28 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 2.1. Mean guard hair diameter (pm) for each of the ten body regions in the three pocket gopher specimens examined. The Duncan post-ANOVA test reveals five groups (designated A-E) within which mean hair diameter was not significantly different. For each region, louse taxa found in greater abundance than expected (Fig. 2.1) are indicated. Mean hair Duncan Greater abundance Region diameter (pm) grouping than expected Lateral nape 45.22 ± 1.53 A Geomydoecus aurei Cheek 44.90 ± 1.76 A Geomydoecus aurei Posterior ventral 42.66 ± 1.28 A B Thomomydoecus minor Anterior ventral 41.40 ± 1.60 AB Thomomydoecus minor Lateral 39.53 ± 1.83 B both species Ventral head 38.72 ± 1.51 B C neither species Dorsal head 38.31 ± 1.54 B C D neither species Nape 35.23 ± 1.53 C D E Geomydoecus aurei Posterior dorsal 34.32 ± 1.67 D E Geomydoecus aurei Rump 33.61 ± 1.75 E neither species 29 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 2.2. Total surface area (cm2) of each pelage region (Fig. 2.1) and number of observed and expected lice per region for the two coexisting species of chewing lice. Data are pooled for the three hosts examined. Thomomydoecus minor Geomydoecus aurei Region Area (cm2) Observed Expected Observed Expected Anterior ventral 58.89 80 59 11 20 Cheek 40.38 2 37 12 14 Dorsal head 35.63 0 35 6 14 Lateral nape 33.01 21 26 29 11 Lateral 56.01 128 50 26 18 Nape 30.01 16 29 14 10 Posterior dorsal 35.25 28 35 62 13 Posterior ventral 67.00 71 62 12 21 Rum p 165.75 153 151 4 54 Ventral head 27.88 17 32 9 10 Total 549.78 516 516 185 185 30 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Therefore, differences are large in both density and distribution of the two louse taxa (Fig. 2.2). DISCUSSION Geomydoecus aurei and Thomomydoecus minor are not evenly distributed throughout the pelage of their host (Thomomys bottae) and show a tendency to subdivide the available habitat dorsoventrally (Figs. 2.1, 2.2). Future studies will determine whether this pattern changes geographically, seasonally, or with the age or reproductive condition of the host. When one considers that lice showed the same distributional pattern on the male and female hosts examined in this study, sex of the host does not appear to influence louse distribution, at least for non- reproductive hosts (as in this study). Given that both genera of lice were found throughout the gopher pelage (with the single exception of the absence of Thomomydoecus in the dorsal-head region; Table 2.2), it is clear that the taxa are interactive (as defined by Brooks 1980) and are able to transit, if not forage and reproduce in, regions of pelage with hairs of different diameters. In fact, regions of high abundance for Geomydoecus aurei encompass almost the entire range of hair diameters (from 34.32 ym to 45.22 ym, Table 2.1). In contrast, regions of high abundance for Thomomydoecus minor include only regions with hairs of intermediate diameter (from 39.53 ym to 42.66 ym, Table 2.1). Despite the general dorsoventral trend in hair diameter (Table 2.1) and a similar dorsoventral trend in louse distribution (Table 2.2; Fig. 2.1), the broad overlap in hair diameters used by the two species (Table 2.1) 31 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. without prohibited reproduction Further owner. copyright the of permission with Reproduced pelage. Asterisks indicate regions that contained more lice than expected in all 3 all in expected than 2.1). Fig. lice 2.1, more (Table ined contained exam that specimens regions gopher indicate pocket Asterisks pelage. Louse density (lice/cm 2) 0 2 1.5 Figure 2.2. Comparative density of lice in the 10 regions of pocket gopher gopher pocket of regions 10 the in lice of density Comparative 2.2. Figure . . 5 5 - - * 03 •-t 0 03 > n O ^ p •n ' o 3 •-* o < 03 Li ' o 3 0 ^ p O rD C/3 o 3 - ^ p *0 < ■-n Li Regions of pocket gopher pocket of Regions -a f* 3 CO ’ o suggests that hair diameter, alone, is insufficient to explain habitat partitioning in this louse community. It is possible, however, that Thomomydoecus lice are less efficient than their competitors at grasping hairs of extremely large or small diameter, yet are superior competitors in regions of intermediate hair diameter. It is also possible that the two louse species differ in their ability to evade grooming pressure from the host, which may vary dorsoventrally. Waage (1979) suggested that areas of overlap in the distribution of co-occurring ectoparasites may receive greater grooming pressure from the host because of higher overall density of parasites. Waage contended that removal of parasites from these areas of overlap (by grooming) would reinforce spatial partitioning of the ectoparasites by removing the ability of parasites to cross corridors between areas of exclusive habitation. However, my observations of captive pocket gophers suggest that they groom only infrequently, and I think it is more likely that the habitat partitioning observed in this study results from differential responses on the part of the two louse species to other microhabitat features such as temperature or humidity gradients or location and density of sebaceous glands of the host (Murray 1957a; 1957b). It seems reasonable to postulate that the dorsal and ventral surfaces of a pocket gopher constitute very different microhabitats, and this hypothesis will be examined in future studies. In addition, future studies of Thomomys bottae populations that host only one of these two species will reveal the extent to 33 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. which competition may be influencing the distribution of these species when they coexist. The fact that hair diameter may have little of no influence on louse distribution at the level of the individual host does not automatically falsify the hypothesis that hair diameter may be an important causal factor influencing louse distribution at higher phylogenetic levels, e.g., among different species and genera of hosts (Reed 1994; Page and Hafner 1996). For example, I have shown (chapter one) dramatic differences in hair diameter among different genera of pocket gophers and have documented a close relationship between hair diameter in the hosts and rostral groove dimensions of their chewing lice. Artificial transfer studies by Reed and Hafner (1997) suggest that lice that normally parasitize species of pocket gophers with narrow hairs may be unable to grasp the wider hairs of larger species of pocket gophers. Finally, studies by Murray (1957a; 1957b) show that hair diameter in sheep may influence ovipositing in their chewing lice. Together, these studies suggest that hair diameter may be a coarse-grained determinant of chewing louse distribution, wherein lice are unable to transfer between hosts with large differences in hair diameter (e.g., Reed and Hafner 1997), but are tolerant of lower levels of variation, such as those observed at the individual and intraspecific host levels. It follows that some other, as yet unknown, environmental parameter, perhaps temperature or humidity, may be the fine-grained determinant of louse distribution at the individual and intraspecific host levels. If so, differential responses to these 34 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. factors by different species of lice may enable stable coexistence of multiple species on a single host individual. 35 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 3 BACTERIAL DIVERSITY IN CHEWING LOUSE HOSTS INTRODUCTION The Hosts: Pocket Gophers of the Rodent Family Geomyidae Pocket gophers (Rodentia: Geomyidae) are fossorial rodents whose geographical range extends from southern Canada through northwestern Colombia. The family consists of six genera and approximately 400 recognized species and subspecies. Evolutionary relationships among genera of pocket gophers are well characterized (Hafner 1982; Hafner and Nadler 1988; 1990; Honeycutt and Williams 1982). These rodents are extremely asocial and live in isolation except during brief mating encounters. Each individual gopher occupies an extensive burrow system that is plugged with soil at all entrances, and populations of gophers are small and isolated from other conspecific populations. Different species of pocket gophers are rarely found in the same region, and then only along narrow zones of parapatry. These natural history characteristics prevent widespread transfer of parasites among individuals in a population, among conspecific populations, or among species. The Parasites: Chewing Lice of the Phthirapteran Family Trichodectidae Chewing lice of the genera Geomydoecus and Thomomydoecus (Phthiraptera: Trichodectidae) are restricted to pocket gopher hosts. 36 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chewing lice are small, hemimetabolous insects no greater than 1 mm in length. They feed on skin detritus and the glandular secretions of pocket gophers (Marshall 1981). Trichodectids are obligate ectoparasites that spend their entire life cycle on the host. The 120 recognized species and subspecies of these two genera are partitioned into four subgenera and 26 species complexes (Hellenthal and Price 1991; 1994). The morphological taxonomy of trichodectid lice has been studied extensively (e.g., Hellenthal and Price 1984; Price et al. 1985). The most recent and comprehensive systematic treatment of trichodectid lice is that of Page et al. (1995), in which 58 morphological characters from adults and first instars were analyzed cladistically. The louse phylogeny generated by Page and his colleagues supports the traditional species complexes proposed by R. Hellenthal and R. Price over the past two decades. This phylogeny also is consistent with molecular-based louse phylogenies proposed by Hafner and Nadler (1988) and Hafner et al. (1994). Previous studies have shown that these lice exhibit a high degree of host specificity (Patton et al. 1984; Timm 1983). Reed and Hafner (1997) showed that chewing lice can survive on non-native hosts, but their ability to colonize new hosts diminishes as the phylogenetic distance increases between the natural host and the surrogate host. Transmission of chewing lice is thought to occur only through direct host- to-host contact (Tim m 1983). H ow ever, D em astes et al. (1998) w ere able to show that transmission was not strictly maternal. The combination of low parasite vagility and obligate contact-transmission of lice limits 37 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. opportunities for colonization of new host species. The aforementioned life-history characteristics of pocket gopher hosts further retard the colonization of new hosts by chewing lice. The absence of widespread transfer of lice among gopher species has, in part, led to the pattern of cophylogeny as documented by Hafner et al. (1994), and results to date suggest that cospeciation is common in this host-parasite assemblage (e.g., Demastes and Hafner 1993; Hafner and Nadler 1988; Hafner et al. 1994; Page 1990b). More vagile species of lice, such as the phthirapteran lice of avian hosts, show more evidence of host-switching and less cospeciation (e.g., Johnson and Clayton in press) than do trichodectid chewing lice associated with pocket gophers. The Endosymbionts: Bacteria The term symbiosis originates from the Greek sym (together) and bios (life). Smith and Douglas (1987) refined this definition and restricted it to include only "long-term associations in close proximity." The definition today encompasses intracellular and extracellular symbionts, and mutualistic as well as parasitic relationships between symbionts. Douglas (1989) reviewed the symbioses of insects and bacteria and determined that approximately 10% of known insect species contain non-parasitic microorganisms. More recently, Werren et al. (1995) estimated that 17% of insect species harbor the bacterium Wolbachia. Modem molecular techniques are demonstrating that Douglas' (1989) figures were vastly underestimated. It is becoming clear that most insects host multiple species 38 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. of microorganisms that are required for specialized functions in the host, such as production of vitamins, synthesis of essential amino acids, digestion of complex foods, and development of offspring during the instar stage. Prior to my research, endosymbiotic bacteria were not known from trichodectid lice of pocket gophers. Ries (1931) used traditional microscopy to search for bacterial endosymbionts in the chewing lice associated with pocket gophers. He was unsuccessful in finding bacteria, although this is possibly an artifact of the microscopy techniques available at the time of study. Eberle and McLean (1983) documented bacterial symbionts in mycetomes of the human body louse (Ped.icu.lus). These bacteria migrate from specialized cells located near the louse midgut to lateral oviducts in adult lice, where they are incorporated into the eggs and passed transovarially to offspring. Saxena and Agarwal (1985) showed that some bird lice harbor endosymbiotic bacteria that may aid in the digestion of keratin-based feathers. My preliminary analyses using PCR, cloning, and sequencing revealed many bacterial taxa associated with trichodectid chewing lice. Although the biological interaction between these symbionts and their host is not yet known, it is likely that one or more of them provide the louse with tissue-degrading enzymes (lice feed on host skin detritus) or essential nutrients (Dadd 1985). None of the louse-associated endosymbiotic bacteria has been cultured outside of its host. 39 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Endosymbiotic Relationships: Insects and Bacteria Numerous relationships have been described between invertebrates and endosymbiotic bacteria (Douglas 1989; Ishikawa 1989; Moran and Telang 1998; Nardon and Grenier 1989; Saffo 1992; for review see Baumann and Moran 1997). One of the best studied endosymbiont systems is that of the gut-inhabiting protozoa and bacteria of termites (Ohkuma and Kudo 1996; for review see Brune and Friedrich 2000). Termites (Reticulitermes: Rhinotermitidae) have an enlarged hindgut (or paunch) that harbors a complex assemblage of microorganisms (Berchtold et al. 1994; Ohkuma and Kudo 1996). This intestinal microflora aids in lignocellulose digestion, methanogenesis, nitrogen fixation, and the maintenance of a low redox potential to prevent entry of foreign bacteria (Ohkuma and Kudo 1996). Wasps of the genus Trichogramma host the cytoplasmically inherited bacterium Wolbachia that usually causes complete parthenogenesis in the host (Schilthuizen and Stouthamer 1997). The intracellular bacteria found in cockroaches (multiple families: Blattaria) are a monophyletic subgroup of the Flavobacter-Bacteroides group that shows evidence of cophylogeny with its host (Bandi et al. 1994). Aksoy et al. (1995) found mycetome-associated endosymbionts in tsetse files (Glossina: Glossinidae) that formed a monophyletic lineage within the y-3 subdivision of the Proteobacteria. Suffice it to say that a wide diversity of bacterial lineages have evolved symbiotic relationships with invertebrates, particularly insects. 40 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The most extensive study of insect-bacterium symbiosis is that of pea aphids (Homoptera: Aphididae) and their endosymbiotic bacteria of the genus Buchnera. Unterman et al. (1989) sequenced the 16S rRNA genes for the primary (p) and secondary (s) endosymbionts of aphids and determined that the p-symbionts were in the gamma subgroup of the Proteobacteria. They determined that the association between the symbionts was ancient, perhaps as old as 420 million years (Myr). Munson et al. (1991) determined that the s-symbionts formed a monophyletic clade in the Enterobacteriaceae. The genus Buchnera was described for the p-symbionts of aphids by M unson et al. (1991). Moran et al. (1993) compared the phylogenetic histories of the endosymbionts and their hosts and determined that there was complete congruence between the phylogenies because of a long history of cospeciation. In addition, Moran et al. (1993) refined the age of the association to 160-280 Myr based on fossil evidence of the host. This study also provided the first convincing evidence of rates of molecular evolution for bacterial lineages (1-2% sequence divergence per Myr). Previously, Ochman and Wilson (1987) proposed a rate of 0.7-0.8% per Myr, but they had no fossil evidence on which to base their calibration. M unson et al. (1993) dem onstrated th at the genom e of Buchnera closely resembles that of its free-living relatives (e.g., E. coli) in size and composition and does not show the reduced genome characteristics of other well-known endosymbiotic bacteria (typically non-essential genes are lost in endosymbiotic bacteria). Subsequent research by Moran and her colleagues 41 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. has focused primarily on rates of evolution in Buchnera. For example, Moran et al. (1995) found that the nucleotide substitution rate was 36 times greater in Buchnera than in its aphid hosts and that Buchnera evolved twice as fast as other closely related bacteria. Brynnel et al. (1998) investigated rates of evolution in the tu f gene of Buchnera to refine the calibration of nucleotide evolution in bacterial lineages. Again, Brynnel and colleagues found that evolutionary rates in Buchnera were faster than those of closely related bacteria. More recently, studies of rates of evolution have been expanded to pairs of closely related sister species outside the aphid system. For example, Ochman et al. (1999) demonstrated that experimental estimates of rates of bacterial evolution (based on laboratory strains) were in conflict with rates estimated from naturally occurring bacteria. Ochman et al. (1999) examined many lineages of bacteria (simultaneously) and found clock-like rates of change when the rates were calibrated with information from fossil pea aphids. Relatively few individual bacteria are passed transovarially from female insects to their offspring. This recurring bottleneck increases the likelihood of endosymbiotic bacteria retaining slightly deleterious mutations. This likelihood is further increased by low population densities attributed to the endosymbiotic lifestyle and the absence of genetic recombination (Buchnera have been shown to have little or no recombination or horizontal gene transfer; Moran and Baumann 1994). "Muller's ratchet" refers to a phenomenon in which slightly deleterious 42 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. mutations gradually accumulate in small, asexually reproducing populations. Moran (1996) described a classic example of Muller's ratchet in Buchnera and used this phenomenon to explain why Buchnera species evolve more rapidly than closely related free-living bacteria. The accumulation of slightly deleterious mutations also was investigated in the ribosomal RNA of Buchnera and other endosymbiotic bacteria by Lambert and Moran (1998). They found that accumulations of these mutations destabilized the secondary structure of ribosomal RNA in all endosymbiont lineages relative to their free-living relatives. Further, Lambert and Moran (1998) demonstrated that this instability has evolved separately in each endosymbiont lineage and thus appears to be a predictable result of the endosymbiotic lifestyle. Baumann et al. (1995; 1997) studied the obligate nature of aphid- endosymbiont associations. They showed that aphids treated with antibiotic agents grow more slowly, show decreased adult weight, and fail to reproduce. All animals require 10 essential amino acids (arginine, histidine, isoleucine, leucine, lysine, methionine, threonine, tryptophan, valine, and phenylalanine), which are usually acquired in the animal's natural diet. Douglas and Prosser (1992) demonstrated that tryptophan was lacking in the pea aphid diet but was produced for them by the endosymbiotic Buchnera. Many other endosymbiotic bacteria overproduce essential amino acids which are lacking in the host's diet (Dadd 1985; Aksoy 1995). Buchnera also shows other features characteristic of the endosymbiotic lifestyle. For 43 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. example, the Buchnera genome contains single copies of genes coding for rRNA (typical of slow growing bacteria) and have high levels of GroEL (a chaperon involved in protein folding) which is commonly found at high levels in endosymbiotic bacteria (Baumann et al. 1997). Clearly, Nancy Moran, Paul and Linda Baumann, and their collaborators have provided the scientific community with a model system for the study of insect-bacterium endosymbioses. Much of the work that lies ahead in the gopher-louse-bacterium system will provide independent tests of their findings and offer additional insight into the evolution of complex endosymbiotic relationships. Environmental Sampling: A Culture-Independent Method The two traditional methods for detection and characterization of endosymbiotic bacteria—light microscopy and culturing—offer little promise in the gopher-louse system. A previous search using traditional light microscopy failed to find endosymbiotic bacteria in the trichodectid lice of pocket gophers (Ries 1931), and most bacteria known today cannot be cultured in the laboratory. Even if successful, both of these methods are likely to underestimate the bacterial diversity found within a host. I used an alternative approach for detection and characterization of endosymbiotic bacteria, an approach known as environmental sampling. This approach uses the intrinsic power of the Polymerase Chain Reaction (PCR) to amplify DNA sequences from bacteria found in an environmental sample. By extracting and selectively amplifying the DNA found in a gram 44 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. of soil, a milliliter of seawater, or an entire chewing louse, one can generate billions of copies of bacterial DNA. This technique has been used extensively in studies of microbial diversity and evolution (Woese 1994; for review see Pace 1997), and it permits culture-independent detection and phylogenetic placement of unknown bacteria (Hugenholtz et al. 1998; W ard et al. 1998). Because this method circumvents the need to culture bacteria in the laboratory, the investigator is no longer constrained to the relatively few bacterial taxa that are cultivable. Although culture-independent methods have broadened our knowledge of bacterial distribution, diversity, and evolution, the method has certain limitations. For example, several studies have shown that chimeric 16S rRNA sequence artifacts can be formed when PCR is used to amplify DNA sequences from mixed populations of bacteria (Komatsoulis and Waterman 1997; Wang and Wang 1997). Suzuki and Giovannoni (1996) showed that certain bacterial taxa were more likely than others to amplify under certain PCR conditions, causing a bias in the kinds and relative numbers of bacteria detected. Additionally, PCR biases can arise merely as the result of differences in 16S rRNA gene copy number (Farrelly et al. 1995; Rainey et al. 1996). T anner et al. (1998) surveyed n e g a tiv e D N A extraction procedures and documented that ambient contamination was easily amplified during PCR with the universal bacteria primers typically used in the culture-independent approach. Tanner and colleagures cautioned that this contamination is unavoidable and recommended use of negative controls to identify potential contaminants. Despite each of these 45 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. legitimate concerns, the culture-independent sampling technique holds great promise for documenting the large number of undescribed bacterial taxa found in environmental samples (Pace 1997). This study used the environmental sampling approach to survey chewing lice for endosymbiotic bacteria. Chewing lice ( Geomydoecus) w ere collected from eight species of pocket gophers in four genera Cratogeomys ( , Geomys, Orthogeomys, an d Thomomys). Because DNA extractions from these specimens contained DNA from both chewing lice and bacteria, primers specific to bacterial 16S rRNA sequences were used to preferentially amplify bacterial DNA. Cloning and cycle sequencing generated DNA sequences for the putative bacterial endosymbionts. These were compared to known bacterial DNA sequences to identify the unknown bacteria. MATERIALS AND METHODS Collection of Specimens Chewing louse samples were collected from their pocket gopher hosts by euthanizing the gopher with chloroform in an airtight canister and then brushing the fur vigorously. All ectoparasites were collected on aluminum foil beneath the pocket gopher host. The louse samples were stored in liquid nitrogen, transferred to the laboratory, and deposited in the LSU Museum of Natural Science Collection of Genetic Resources. The following lice were collected: Geomydoecus oklahomensis from Geomys bursarius halli (host number LSUMZ 31463); and G. b. major (LSUMZ 31448); Geomydoecus scleritus from Geomys pinetus; 46 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Geomydoecus expansus from Cratogeomys castanops (LSUMZ 31455); Geomydoecus panamensis from Orthogeomys cavator (LSUMZ 29253); Geomydoecus setzeri from O. cherriei (LSUMZ 29539); Geomydoecus costaricensis from O. heterodus (LSUMZ 29501); Geomydoecus chapini fro m O. hispidus (LSUMZ 29232); Geomydoecus setzeri from O. underwoodi (LSUMZ 29493); and Geomydoecus centralis from Thojjiomys bottae (LSUMZ 29569). Extraction Protocol Adult chewing lice were washed twice in a solution of 400/d saline EDTA buffer (containing 150mM NaCl, lOmM EDTA, pH=8.0), 10/d of 25% SDS (sodium dodecyl sulfate), and 5/d of lOmg/ml lysozyme to remove bacteria from the outer surfaces. The lice were then placed in 1.5ml microcentrifuge tubes along with 400/d of saline EDTA buffer, and 5/tl lysozyme (lOmg/ml) and were crushed with sterile micropestles. Two negative controls were also used; one contained all the reagents used in the DNA extraction process and the other contained the same reagents plus a micropestle to insure that nothing was contaminated with bacteria. The extraction solution was incubated at 37°C for 30 minutes. Five microliters of proteinase-K (15mg/ml) and 10/d of SDS (25%) were added, and the solution was incubated at 55°C for 30 minutes. Genomic DNA was extracted by adding 400/d of phenol/chloroform, vortexing, and spinning in a microcentrifuge for two minutes at 14,000 rpm. The supernatant, which contained the extracted DNA, was carefully 47 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. removed, and the extraction procedure was repeated two more times. One milliliter of 95% ethanol (EtOH) was added, and the solution was incubated on ice for 24 hours to precipitate the DNA. Centrifugation at 14,000 rpm for 20 minutes concentrated the DNA into a pellet. The EtOH was decanted, and the microcentrifuge tubes were dried in a vacuum centrifuge for 15 minutes. The resulting pellets were resuspended in 50/d of TE buffer. Polymerase Chain Reaction The polymerase chain reaction was used to amplify copies of bacterial 16S rRNA from genomic DNA. Amplifications were performed using primers 27-f (5'-GAG TTT GAT CCT GGC TCA G-3') and 1525-r (5'-AGA AAG GAG GTG ATC CAG 00-3') which were designed for their ability to anneal to many types of bacteria. Genomic DNA (2 /d), 3/d of each primer (10/tM), 3/d of deoxynucleoside-triphosphate (dNTP) mixture (dATP, dGTP, dCTP, and dTTP, each 1/iM), 3/d of MgCl2 (25mM), and 1 unit of Taq D N A polymerase were combined in a 50/d PCR reaction. Negative PCR controls, which contained no DNA template, were used to test for contamination of the PCR reagents. Thirty-five thermal cycles were performed, each with a 1 minute denaturation period of 94°C, a 1 minute annealing period of 56°C, followed by a 1 minute extension period of 72°C. After 35 cycles, a single extension time of 10 minutes at 72°C was used to facilitate polymerase activity and extend PCR products. Amplification products were visualized with ethidium bromide after electrophoresis in a 1% agarose gel. Samples were electrophoresed adjacent to a DNA ladder to determine the fragment 48 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. size of the amplified DNA. Reactions containing fragments of the appropriate size (ca. 1,530 base pairs "bp") were cleaned with Qiaquick™ spin column PCR purification kits (Qiagen®, Valencia, CA) as prescribed by the manufacturer. Cloning Cleaned PCR fragments were ligated into pCR®4-TOPO® plasmid vectors (Invitrogen® TOPO TA Cloning® Kit for Sequencing, Carlsbad, CA). Ligations were performed at room temperature with topoisomerase as prescribed by the manufacturer. Ligated plasmids containing PCR fragment inserts were transformed into TOPIO One Shot® competent £.coli cells as prescribed by the manufacturer. They were incubated at 37°C for 30 m inutes and plated in 50/zl, 100/d, and 250/d aliquots onto ampicillin-resistant (50/ig/ml) Luria broth (LB) agar plates. The plates were incubated overnight at 37°C. Bacterial colonies were picked the next day based on either blue/white screening or presence/absence screening for clones that contained a plasmid insert. Positive clones were collected and transferred to 5ml culture tubes containing LB broth and 50/ig/ml ampicillin. The cultures were shaken overnight at 37°C. Two methods were used to isolate cloned plasmid DNA from the genomic DNA of £. coli. The first method used S.N.A.P. mini-prep kits (Invitrogen®, Carlsbad, CA) to separate plasmid DNA from £.coli genom ic 49 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. DNA by centrifugation. The second method employed an additional PCR step using primers located in the plasmid (M13f: 5'-GTA AAA CGA CGG CCA G-3' and M13r: 5'-CAG GAA ACA GCT ATG AC-3'). These primers flanked the DNA template insertion site and matched the plasmid sequence exactly. Both procedures produced high quality PCR fragments. Cycle Sequencing and BLAST Searches Cycle sequencing was performed on cloned plasmid vectors after mini-preparations or after reamplification with plasmid primers. Both procedures generated highly accurate and unambiguous DNA sequences. Cycle sequencing was performed by using the ABI PRISM™ Big Dye™ kit (PE Applied Biosystems, Foster City, CA) as prescribed by the manufacturer, with the following exceptions. The "ready reaction mix" provided in the kit was diluted with 2.5X sequencing buffer (200mM Tris, 5mM MgCl2, pH=9.0). The reactions contained 2/zl "ready reaction mix" (not the prescribed 8/d), 2.8/d of sequencing buffer (2.5X), 3.2/d of primer (.5/iM), and 2/d of DNA template. The thermal cycling protocol consisted of a denaturation phase of 96° for 10 sec, followed by an annealing phase of 48-50° for 5 sec, followed by an extension phase of 60° for 4 min. The cycling protocol was repeated a total of 25 times. This system uses dye-labeled terminators that fluoresce upon laser contact during electrophoresis on an ABI 377-XL automated sequencer. Two plasmid primers (M13f and M13r) and four internal 16S rRNA primers (536f: 5'-CAG CMG CCG CGG TAA TWC-3', 1114f: 5'-GCA ACG 50 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. AGC GCA ACC C-3', 519r: 5'-GWA TTA CCG CGG CKG CTG-3', and 960r: 5'- GCT TGT GCG GGY CCC CG-3') were used. Cycle sequencing products were cleaned with the ethanol/sodium acetate procedure outlined in the ABI Big Dye manual. Automated DNA sequencing was performed with an ABI™ 377-XL DNA Sequencer and provided sequences approximately 700 bp in length. Sequencing reactions using primers M13f and M13r generated approximately 1,000 bp of the 16S rRNA gene (base pairs 1-500 and 1,050- 1,550). These fragments were compared to the National Center for Biotechnology Information database (Genbank) by using BLAST searches. Complete 16S rRNA sequences (-1,550 bp) were required for phylogenetic study of several bacterial lineages. Therefore, some clones were sequenced in both forward and reverse directions by using the 6 primers listed above. The computer program Sequencher® 3.1 (Gene Codes Corporation, Ann Arbor, MI) was used to proofread and join contiguous fragments of DNA sequence into a single consensus sequence for each cloned sample. Completely sequenced clones with similar identifications could then be aligned with Sequencher 3.1. Divergent sequences were aligned by using ClustalX (Thompson et al. 1997). BLAST searches were used to query GenBank (NCBI National Center for Biotechnology Information, http://www.ncbi.nlm.nih.gov) for DNA sequences that showed high sequence similarity to the unknown bacterial DNA sequences amplified from chewing louse extractions. The results were 51 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. returned via e-mail or directly over the World Wide Web. The unknown and known sequences (called high-scoring segment pairs, or HSPs) are aligned and compared statistically at NCBI. The distribution of a randomly chosen set of HSPs follows a Poisson distribution. The BLAST algorithm uses this distribution to determine the probability associated with the similarity between HSPs. The BLAST program reports both probability values (P-values) and scaled E-values, the latter being easier to interpret. The E-value decreases as sequence similarity increases. For example, E- values of 5 and 10 have corresponding P-values of 0.993 and 0.99995. However, when E < 0.01, P-values and E-value are nearly identical. The BLAST program ranks and lists all HSPs by E-values. The top BLAST result (if significant at E < 0.01) was used as temporary identification for the unknown bacterium until further study could reveal its accurate taxonomic placement. The BLAST program also provides pairwise alignments of HSPs, values of percent sequence similarity, and information regarding gaps in the alignment of the two sequences. RESULTS BLAST searches of 339 16S rRNA sequences (-500 bp each) identified 35 lineages of bacteria in eight divergent groups of the domain Eubacteria (Table 3.1). Many of the searches found no significantly similar DNA sequences in GenBank and thus, are not listed in Table 3.1 (n=112/339). These could represent PCR artifacts, chimeric sequences of multiple organisms (Komatsoulis and Waterman 1997; Robison-Cox et al. 1995; 52 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 3.1. Bacteria associated with chewing lice identified to the level of genus by means of BLAST searches of GenBank. Types of Bacteria Genus Habitat Thermus/Deinococcus group • T herm us g ro u p Meiothermus Proteobacteria • Alpha Proteobacteria Rhizobiaceae Ensifer Methylobacterium P l a n t s Air conditioning units Rhodospirillaceae Phaeospirillum Rhodospirillum D e a d . S ea Hot springs W aste water Soil and salt marsh Sphingomonas group Sphingomonas Plant roots S o il River sediments • Beta Proteobacteria Burkholderia group Burkholderia Soil OEurope & Japan) Cystic Fibrosis patients Rotting bark Plant roots Blood cultures Rhizobacterium S o il Plant roots Comamonadaceae Acidovorax S o il Activated sludge Leptothrix Activated sludge Alcaligenaceae Alcaligenes W hirlpool baths Bone marrow Cerebrospinal fluid S o il Activated sludge P l a n t s Rhodocyclus Azoarcus Ultramicrobacterium group Ultramicrobacterium Gamma Proteobacteria Psuedomonas group Pseudomonas Plant roots Oil degredation S o il M u sh ro o m s Activated sludge Groundwater sources (table cont.) 53 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Types of Bacteria Genus Habitat • Gamma Proteobacteria (cont.) Psuedomonas group (cont) Acinetobactev Cerebrospinal fluid Blood, urine, pus Nose, throat, vagina M o u se sk in S o il Activated sludge S e a w a t e r Legionellaceae Legionella Hospital patients Xanthomonas group Xanthomonas Hospital respirator Human urine Various plants • Delta Proteobacteria Desulfuromonas group Geobacter S o il F irm ic u te s • Actinobacteria Actinomycetales Brevibacterium Oil degredation Cheese production Gordonia S o il Aquatic sediments Biofilters (waste treatment) Janibacter — Kitasatospora — Micrococcus S o il Oil degredation S k in Nocardioides P la n ts S o il S e a w a t e r Contaminated ground water Propionibacterium S k in Swiss cheese • Bacillus/Clostridium group Bacillus/Staph, group Facklamia Blood, urine, etc. Powdered tobacco (snuff) Staphylococcus S k in Hymenoptera in amber Clostridiaceae Peptostreptococcus H u m a n g u t m icrobe Streptococcaceae Streptococcaceae Throat, vagina Vertebrate lungs A v ia n c ro p Verrucomicrobia • Verrucomicrobiales Verrucomicrobiaceae Verrucomicrobium — Planctomycetales Planctomyces S p o n g e Undescribed Eubacteria Undescribed taxa Various habitats 54 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Suzuki and Giovannoni 1996; Wang and Wang 1997), sequences for species of bacteria not yet archived in GenBank, or fragments that failed to sequence correctly. Some searches found significantly similar sequences in GenBank that were themselves labeled "unknown" (e.g., "unknown Eubacterium" and "unknown Proteobacterium"). These "unknown" sequences represent species that have yet to be described. The louse-associated sequences often showed high sequence similarity with the "unknown" sequences in G enB ank. Table 3.2 shows pocket gopher and chewing louse hosts, their general collection locality, and the number and type of bacteria found in DNA extracts from chewing lice. Many of the 23 lineages of bacteria that were found in only one or two louse species (e.g., Acidovorax, Brevibacterium, and Ensifer) likely represent transient bacteria acquired through the louse diet. Twelve bacterial lineages (Table 3.2) were found in three or more species of chewing lice and warrant further investigation. However the "unknown Proteobacterium" and "unknown Eubacterium" categories (Table 3.2) likely contain several distantly related species of bacteria and cannot be considered single "lineages." Six lineages (e.g., Alcaligenes, Azoarcus, Burkholderia, Methylobacterium, Propionibacterium, and Staphylococcus) were found in four or more DNA extracts from chewing lice. Staphylococcus-like clones had the highest prevalence and were found in seven of eight louse species (Table 3.2). Many Staphylococcus spp. are skin-associated bacteria (Table 3.1) so it is not surprising to find them 55 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 3 1 1 1 1 2 1 of Hosts Number 1 1 1 1 1 1 1 1 1 1 2 1 6 4 1 7 4 14 3 of 20 4 22 6 (table cont.) Clones Number 1 1 1 minor bottae centralis Thom om ys New Mexico Geomydoecus Thomomydoecus 2 heterodus Costa Rica costariccnsis Geomydoecus Orthogcomys 1 7 3 setzeri Costa Rica undcrwoodi Geomydoecus Orthogcomys 1 3 1 5 9 1 2 3 cavator Costa Rica panam ensis Geomydoecus Orthogcomys l Florida Geomys pinetus scleritus Geomydoecus 1 1 2 3 9 Geomys Missouri bursarius geom ydis Geomydoecus 1 1 1 1 1 1 4 2 5 8 expansus castanops New Mexico Geomydoecus Cratogeomys Table 3.2. Table 3.2. Endosymbiotic bacteria identified from nine taxa of chewing lice by means of environmental extraction, Kilasalosporia Legionella Leplothrix Lnicmioslic Gardania Geobacter Brcvibacterium Burkholderia Ensifer Janibacter Facklamia Azoarcus Locality: Acidovorax Aciin'tobacler Alcaligenes Host: Chewing Louse Gopher Host: PCR, PCR, sequencing, and searching GenBank databases via nucleotide BLAST recognition search protocol. "Unk." = unknown. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 1 1 1 1 1 4 4 5 7 1 3 3 2 3 5 1 2 2 1 2 2 5 1 1 1 1 1 7 5 1 2 9 2 2 1 1 2 2 24 16 43 13 227 17 24 29 329 21 45 13 48 46 1 2 1 1 8 9 1 1 l 1 1 10 19 16 35 1 1 9 3 4 1 51 17 1 1 3 1 1 2 2 l 2 3 4 5 1 1 16 18 77 78 l 48 Bacterial Types Identifiable Verrucomicrobium Examined Clones Xantlwimmas Total Clones llllramicrolHictcrium Unk. ProteobacteriumUnk. Eubacterium 4 Staphylococcus Strqitococcus Rlmlospirillum Sphignomonas Pseudomonas Rliizobactcrium Planctomyces Propionibaclerium Phacospirillum Peptostreptococcus Nocardioides O c h ro M u m Mciallienmis Micrococcus Melhylobacterium Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. associated with insects that feed on skin detritus. Representatives of three bacterial genera were found in five {Burkholderia) or six {Methylobacterium, and Propionibacterium) species of chewing lice. Some species ofBurkholderia and Methylobacterium are known plant associates (Table 3.1), so the clones found in the chewing louse samples could be transient bacteria ingested by the lice. Bacteria associated with plants (especially roots) are likely to come in contact with lice that live in the fur of subterranean, herbivorous rodents. Species in the genera Acinetobacter, Legionella, Pseudomonas, and Xanthomonas are closely related members of the gamma subgroup of the Proteobacteria. Many of the endosymbiotic bacteria of insects are found in this subgroup of Proteobacteria. One species of chewing louse {Geomydoecus expansus) hosted eight of the most common types of bacteria (Table 3.2). The absence of one of the common bacterial species {Azoarcus) in G . expansus may represent incomplete taxon sampling. Geomydoecus centralis and Thomomydoecus minor (collected from Thomomys bottae) hosted a combined total of only six types of bacteria. The predominance of only a few species could be caused by PCR bias, local extinction of other bacteria, or incomplete taxon sampling. In total, 227 of 329 clones were tentatively identified as one of 35 types of bacteria. DISCUSSION Sampling Strategy Ribosomal DNA sequences from 35 types of bacteria were amplified and sequenced from DNA extracted from chewing louse samples. Certain types of bacteria 58 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. could be found in most of the chewing louse hosts, whereas others were found only in a single host (Table 3.2). Bacteria detected in only a few hosts could, in fact, be associated with all eight louse species but may have been missed in my study due to incomplete taxon sampling. This is a likely possibility, considering that approach was intentionally wide in scope and relatively shallow in depth. An exhaustive survey, including the sequencing of more clones from each DNA extraction, might help to reduce taxon sampling bias. To test this, I plotted the number of different types of bacteria found in my study against the total number of clones sequenced, which varied from one chewing louse extraction to another (Fig. 3.1). Although Figure 3.1 shows that more taxa are detected when more clones are sequenced, the low slope of the line (a = 0.1844) indicates that the discovery of additional taxa will require examination of very large numbers of additional clones. If I had sampled all of the chewing louse extracts exhaustively, then the number of bacterial taxa detected would have had no relationship with the number of clones sequenced. This pattern clearly demonstrates that the sampling effort in this study did not detect all of the bacterial lineages in chewing louse samples. I used rarefaction curves to obtain rough estimates of the total number of clones required to adequately sample bacterial taxa in two chewing louse hosts. The rarefaction curves tally the number of new taxa of bacteria detected (Y-axis) as one increases the number of clones examined (X-axis). For example, if a sampling 59 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 50 'O au > 40 ■ © 4> a >> 30 . H ’E (j « 20 ■ © I I ■ i ■ ■ - t 10 20 30 40 50 60 70 80 90 Total Number of Clones Sequenced Figure 3.1. Regression of the number of types of bacteria found as a function of the total number of clones sequenced for each of the seven gopher-louse host pairs listed in Table 3.2. 60 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. effort of 77 (Fig. 3.2) or 78 (Fig. 3.3) clones was sufficient to detect all (or nearly all) bacterial lineages found in association with a chewing louse host, one would expect to see rarefaction curves in Figs. 3.2 and 3.3 gradually become asymptotic to some value on the Y-axis. However, Figs. 3.2 and 3.3 instead show that new bacteria are being detected even as the last few clones are sequenced. It would appear that many additional clones will need to be examined before concluding that I have an adequate representation of the bacterial community associated with chewing lice. Polymerase Chain Reaction Bias It is important to note that the presence of certain bacteria in an experimental extraction may affect the ability to amplify the DNA of other bacteria. It is not possible to assess bacterial species abundance by merely counting the number of clones of a particular bacterium because it has been shown that bacterial DNA sequences occurring in low abundance in an extract can amplify more rapidly than more abundant DNA sequences if the PCR conditions are better suited to the rarer sequence (Suzuki and Giovannoni 1996). Information on presence or absence of bacteria amplified from mixed-template PCR may be biased in the same way. Although use of primers from conserved regions of the 16S rRNA permits studies of bacterial diversity, conserved primers also introduce biases. I suggest that in future studies, conserved primers should be used initially to detect and identify novel types of bacteria. However, after the initial 61 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 30 Total Number of Clones Sequenced Figure 3.2. Rarefaction curve showing the number of clones identified from Geomydoecus expansus compared to the total number of clones surveyed. 62 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. without prohibited reproduction Further owner. copyright the of permission with Reproduced G eomydoecus geomydis eomydoecus G Types of Bacteria Identified iue33 Rrfcincreso n h u e fcoe dniidfo from identified of clones ber num the ing show curve Rarefaction 3.3. Figure 10 20 30 • ■ 10 Total Number of Clones Sequenced Clones of Number Total compared to the total num ber o f clones surveyed. clones f o ber num total the to compared 20 30 40 63 50 070 60 090 80 survey, primers specific to target bacterial taxa should be used, especially if the researcher intends to search for those bacteria in multiple hosts. This procedure may help to avoid some of the problems associated with mixed-template PCR. The Ubiquity of Bacteria Of the 35 types of bacteria found in this study, eight are known to occur in association with plants and 12 have been recovered from soil samples (Table 3.1). Many species of bacteria are thought to be ubiquitous species worldwide. Six genera detected in chewing louse extracts (Acinetobacter, Acidovorax, Alcaligenes, Comamonas, Leptothrix, and Pseudomonas) also were found in a study of activated sludge from a waste water treatment plant. Other bacteria amplified from chewing lice have been detected in seawater, deep sea sediments (including the Marianas trench), hot springs, and desert soils. Considering this remarkable diversity of habitats, it is reasonable to assume that certain species of bacteria detected in this study may be geographically widespread, if not ubiquitous. The fact that many of these geographically and ecologically distinct organisms have nearly identical 16S rDNA sequences is even more remarkable. Perhaps the best evidence of ubiquity in bacteria is reviewed by Staley and Gosink (1999). They are interested in psychrophilic, free-living bacteria that inhabit polar sea ice. These bacteria live on opposite poles of the earth and cannot live in the intervening warm tropical waters. Despite the lack of an apparent means of genetic exchange, their results suggest that there has been recent gene flow 64 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. between the two polar ice cap populations of bacteria. In short, we know that some species of bacteria show evidence of gene flow over incredible distances, whereas others (e.g., the endosymbiotic Buchnera associated with pea aphids) have virtually no means of genetic exchange over comparatively short distances. The Endosymbionts of Trichodectid Chewing Lice I have documented that there are at least 35 lineages of bacteria associated with the pocket gopher and chewing louse system. However, many of these are likely soil and plant microbes that are transients ingested by chewing lice. Some may be skin bacteria associated with the louse's mammalian host, whereas others may be a part of the natural gut flora of the chewing louse. Whether any of these bacteria are truly endosymbiotic—as are those found in mycetomes of other insects—remains untested. To document that bacteria are truly endosymbiotic, one must: 1) find the endosymbiont in every host; 2) fail to find the endosymbiont in the environment as a free-living bacterium; 3) use in situ hybridization to probe the host for DNA sequences unique to the putative endosymbiont to document that it is living within the host; 4) remove the endosymbiont and document negative effects on the host; and 5) re-infect these bacteria-free hosts with the endosymbiont and show reversal of the negative effects;. These tests will document that the symbiont resides in the host and is an integral part of the hosts existence. 65 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. It is beyond the scope of this dissertation to determine whether the 35 lineages of bacteria found in chewing lice are, indeed, true endosymbionts. However, an evaluation of bacterial sequence similarity and phylogeny may be used to infer endosymbiosis. For example, if a lineage of bacteria shows phylogenetic structuring that matches that of its host, the most probable cause is cospeciation. Cospeciation, in turn, suggests a long history of association between the bacteria and its host. In the present study, it is possible that a lineage of bacteria found in chewing lice is actually cospeciating with the pocket gopher and is only a transient in the gut of the louse. It is conceivable that one of the skin-associated bacteria of pocket gophers has cospeciated with its host. If so, this bacterial lineage likely would show evidence of cophylogeny with the chewing louse lineage, given that chewing lice are known to have cospeciated with their pocket gopher hosts (Hafher et al. 1994) as well. In an effort to determine whether any of the above lineages show evidence of cospeciation, I sequenced the entire 16S rRNA gene for clones identified as Acinetobacter, Legionella, Pseudomonas, Staphylococcus, an d Xanthomonas (chapter 5). These species were selected for the initial analysis because Staphylococcus was detected in a wide diversity of hosts (Table 3.2), and the other lineages are part of the gamma subgroup of Proteobacteria that contains many endosymbiotic bacteria. Several other lineages detected in chewing lice also warrant further study (e.g., Burkholderia an d Propionibacterium), but were not investigated here. 66 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 4 BACTERIAL DIVERSITY IN CHEWING LOUSE EGGS INTRODUCTION Most insects host internal bacteria, some of which are transient visitors, whereas others are long-term endosymbionts. There are numerous well-described insect-bacteria associations in which the endosymbiotic bacteria are required for the survival of their host (Douglas 1989). Pea aphids, for example, require an endosymbiotic bacterium (Buchner a) to reproduce, and they fail to produce offspring if they are treated with antibiotics that kill their endosymbionts (Douglas 1989; Ishikawa 1989). Termites must have a large assemblage of microorganisms to aid in the digestion of cellulose (Douglas, 1989). It is often crucial for insects to have the required complement of endosymbionts immediately upon emerging from the egg (e.g., when the bacteria provide essential nutrients). Thus, bacteria are transmitted transovarially in pea aphids (Douglas 1989; Ishikawa 1989), tsetse flies (Aksoy et al. 1997), hydrothermal vent invertebrates (Cary et al. 1993), cockroaches (Sacchi et al. 1988), leafhoppers (Purcell et al. 1986), human body lice (Eberle and McLean 1983), Phthirapteran bird lice (Saxena and Agarwal 1985), and many other invertebrates. In some insects (e.g. cockroaches, ants, and certain beetles) the endosymbionts migrate to the host ovarioles where they coat the outer surface of developing eggs. Some 67 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. hatching insects eat their own egg case as a source of energy and thus, are inoculated with necessary microorganisms (Douglas 1989). In other insects, the endosymbionts migrate to reproductive structures and actually become incorporated inside the developing egg, which insures transmission to the offspring (Douglas 1989). In human body lice ( Pediculus), mycetome-associated bacteria leave their specialized cells via a single opening and migrate along the digestive tract to the oviduct Eberle and McLean (1982; 1983). The endosymbionts migrate directly to the oviducts, even when the oviducts have been experimentally moved to unnatural locations within the insect. In weevils, one species of endosymbiont is permanently associated with the host's oocytes (reviewed by Douglas 1989). These symbionts are present in mature oocytes, but leave to infect developing oocytes in adult weevils. Although these endosymbionts are passed through the eggs to offspring of both sexes, they eventually are extirpated in the males. There are also mycetome- associated endosymbionts in weevils that have a more typical mode of vertical transmission in which the symbionts migrate from their specialized cells to the ovaries to be incorporated into the eggs. Thus, weevils are host to at least two distinct lineages of endosymbionts that have evolved different mechanisms of vertical transmission. The complex mechanisms involved in transovarial transmission of endosymbionts doubtlessly evolved over countless generations. As a result of this intimate, long-term relationship between these highly specialized 68 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. host-parasite associates, many of them also show evidence of cophylogenv. I am interested to know if the eggs of chewing lice contain bacteria that are transmitted transovarially. If so, these bacteria are almost certainly long term endosymbionts, and are therefore good candidates for the study of cophylogeny. Although Saxena and Agarwal (1985) documented the presence of mycetocyte-associated bacteria in the ovaries of some Phthirapteran bird lice, and Saxena et al. (1985) discussed the history of bacteria associated with the ovaries of many species of lice (including mammal lice in the family Trichodectidae), no study has determined whether mycetocyte-associated bacteria are present in the trichodectid lice that live on pocket gophers. In this study, I used PCR, cloning, and DNA sequencing to search for 16S rRNA sequences from bacteria associated with louse eggs. MATERIALS AND METHODS Twenty chewing louse eggs from Geomydoecus eioingi were collected from the fur of their host, Geomys bursarius, and deposited into each of two 1.5ml microcentrifuge tubes. Each tube of 20 lice was then washed four times in a solution containing 400/rl saline EDTA buffer (containing 150mM NaCl, lOmM EDTA, pH=8.0), 5/d lysozyme (lOmg/ml), and 10/d SDS (25%). This treatment dislodged and removed bacteria from the outer surface of the egg. The solution used to wash the egg was removed after each treatment and transferred to a separate sterile microcentrifuge tube; the louse eggs remained in the same 1.5ml microcentrifuge tube throughout the entire 69 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. process. Extreme care was taken not to rupture the eggs or accidentally remove them during the wash cycles. After the washes, the eggs were crushed with a micropestle. Two negative controls were used during DNA extraction. The first contained all of the reagents used in the DNA extraction process, and the second contained the same reagents plus a micropestle (identical to the one used to crush the eggs in the experimental samples) to insure that none of the extraction reagents or tools was contaminated with bacteria. Each of the four wash solutions, the controls, and the eggs were then digested with proteinase K as described in the previous chapter. Genomic DNA was then extracted by adding 400/d of phenol/chloroform, vortexing, and spinning in a micro centrifuge for two minutes at 14,000 rpm. The supernatant, which contains the extracted DNA, was carefully removed and the extraction procedure was repeated two more times. One milliliter of 95% EtOH was added, and the solution was incubated on ice overnight to precipitate genomic DNA. Centrifugation at 14,000 rpm for 20 minutes concentrated the DNA into pellets. The EtOH was decanted, and the microcentrifuge tubes were dried in a vacuum centrifuge for 15 minutes. The resulting pellets were resuspended in 30/d of TE buffer. The PCR protocol and primers (27-f and 1525-r) described in the previous chapter were used to amplify bacterial 16S rRNA sequences from the wash solutions and the louse eggs. Cloning and DNA sequencing also followed the procedure outlined in the previous chapter. BLAST searches 70 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. in GenBank provided tentative identifications for unknown bacterial sam ples. RESULTS AND DISCUSSION The PCR reactions produced bright, distinct DNA bands for both the washes and the eggs, whereas the negative controls only produced faint smears of DNA that could not be cloned. I judged that this slight contamination was insufficient to cause spurious PCR amplifications in the experimental extracts. Seventy-four clones were sequenced from environmental samples of chewing louse eggs. Most of the bacterial clones from the eggs (50 of 74) were bacteria in the genus Acinetobacter, and Staphylococcus-like bacteria were found in 16 of 74 clones (Table 4.1). A third lineage in the genus Pseudomonas was found in 8 of 74 clones. Sixty-five clones were sequenced from the wash solutions used to remove bacteria from the outside of the eggs. Acinetobacter accounted for 23 of these 65 clones, Staphylococcus 22 of the 65 clones, an d Afipia, Pseudomonas, and Dyadobacterium were found in a few clones each (Table 4.1). From these data it appears that several types of bacteria (notably Acinetobacter, Afipia, Dyadobacterium, Pseudomonas, and Staphylococcus) are present on the outer surface of chewing louse eggs. If I assume that my washing technique effectively removed the external bacteria from the eggs, then Acinetobacter, Staphylococcus, and Pseudomonas likely occur inside chewing louse eggs. It is also possible, however, that the washing protocol 71 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 4.1. Types of bacteria found in DNA extracts of chewing louse eggs. Bacteria were identified, by BLAST searches of the Genbank DNA sequence database. Bacteria Found Egg1 Egg 2 Wash 1 Wash 3 Acinetobacter 18 32 16 7 Afipia 0 0 10 0 Dyadobacterium 0 0 4 1 Pseudomonas 5 3 4 1 Staphylococcus 15 1 0 22 Total 38 36 34 31 72 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. did not remove all bacteria from the outer surface of the egg. If so, then Acinetobacteria, Staphylococcus, and Pseudomonas must adhere strongly to the outer surface of louse eggs, as these taxa were the last removed by washing. Results of the preceding analysis cannot be used to test these two hypotheses, so I designed another protocol to search for bacterial DNA inside louse eggs. Twenty louse eggs were washed as before, but this time in a series of 10 washes. The first, fifth, and tenth washes were amplified by using PCR, as were the eggs themselves and the appropriate controls. All three washes produced PCR products of the target length (ca. 1,530 bp), just as in the previous analysis. However, the extract containing only chewing louse eggs repeatedly failed to produce PCR bands (despite the use of numerous PCR protocols), which suggested the absence of bacterial DNA within the egg. To determine that the DNA extraction protocol was effective, I amplified the egg extracts using primers that anneal to the cytochrome b gene of chewing lice (obviously chewing louse eggs should contain chewing louse DNA). Certainly there is more host (chewing louse) DNA in a louse egg than endosymbiont DNA. These PCR reactions produced single, bright DNA bands that, when sequenced, were found to be louse cytochrome b sequences. Although this protocol does not preclude a methodological problem with bacterial extractions from chewing louse eggs (discussed below), it is consistent with the hypothesis that the chewing louse eggs contain no internal bacterial DNA. 73 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. A constant concern in any attempt to amplify endosymbiont DNA is the far greater concentration of host DNA compared to that of the endosymbiont. The resulting swamping effect can be overcome during PCR by use of highly specific DNA primers that preferentially anneal to bacterial DNA. However, when PCR reactions fail, one cannot rule out low DNA copy number as the cause of the failure. I tried to quantify the amount of DNA in each of the extractions in this study, including the washes that amplified successfully, but all samples contained DNA in such low quantities that they failed to register when examined with a fluorometer. It is conceivable that the failure of PCR reactions containing egg extracts is the result of either no DNA template (i.e., sterile eggs) or the result of low copy number of bacterial DNA. It is not possible to distinguish between these two possibilities at this time. We know from other studies that the number of bacteria transferred through the egg is always low (Douglas, 1989), so the hypothesis of low copy number seems likely. However, we also know that in the absence of a long-term endosymbiosis with transovarially inherited bacteria, the eggs of most animals are, in fact, sterile at the time of deposition. Endosymbionts are sometimes transmitted vertically among insect hosts by adhering to the outer surface of the insect egg (Douglas 1989). This mechanism for vertical transmission of endosymbionts may be evolutionarily intermediate between (primitive) contact transmission and (derived) transovarial transmission. My results are consistent with the 74 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. hypothesis that bacteria are found on the outer surface of chewing louse eggs, but not within. However, conclusive evidence is lacking because my tests cannot distinguish between bacteria that adhered to the egg's surface while the egg was inside the louse versus bacteria that adhered to the egg's surface after the egg was laid. It is interesting to note that each of the taxa of bacteria found in association with chewing louse eggs also was found in abundant numbers in the louse extracts (see previous chapter). Further analyses using microscopy and in situ hybridization may reveal whether the eggs contain, or are coated with, truly endosymbiotic bacteria. In situ hybridization also could be used to identify bacteria associated with the ovarioles of chewing lice to further understand vertical transmission of these bacteria. 75 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTERS STUDIES OF COPHYLOGENY: CHEWING LICE AND ENDOSYMBIOTIC BACTERIA INTRODUCTION Host-parasite systems are intrinsically interesting to evolutionary biologists because they often represent long and intimate associations between two or more groups of organisms that are distantly related and quite dissimilar biologically. This long history of association often leads to reciprocal adaptations in the hosts and their parasites (classical coevolution or coadaptation) as well as contemporaneous cladogenic events in the two lineages (cospeciation or cophylogeny). Comparative phylogeneticists are particularly interested in the phenomenon of cophylogeny because cospeciation events identify temporal links between the host and parasite phylogenies, and thus provide an internal time calibration for comparative studies of rates of evolution in the two groups. Hafner and Nadler (1990) proposed a protocol for investigation of cophylogeny that was designed to remedy many of the problems that hampered earlier studies. They argued that the minimal requirements for a valid test of cophylogeny were: 1) independent host and parasite phylogenies; 2) well-corroborated phylogenies based on tree-building algorithms with explicit assumptions; and 3) rigorous statistical tests of similarity between the host and parasite phylogenies. In addition to these 76 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. basic requirements, they emphasized that a study of cophylogeny would be enhanced significantly by comparative investigation of molecular (as opposed to morphological) differentiation in the hosts and their parasites. This framework for investigation of cophylogeny has been used in numerous published studies, and the number of molecular-based studies of cophylogeny has increased rapidly (Page in press). Previous studies of host-endosymbiont cophylogeny, particularly those involving aphids and their bacterial endosymbionts (Moran et al. 1993; 1995; Munson et al. 1991; Unterman et al. 1989), have shown remarkable levels of concordance between host and symbiont phylogenies. Using host divergence dates estimated from fossil evidence, Moran et al. (1993) were able to infer bacterial divergence dates in the absence of a fossil record. This inference would not have been valid without firm evidence of cophylogeny. Similarly, Moran et al. (1995) used cophylogeny as a framework to compare relative rates of molecular change in aphids and their endosymbiotic bacteria, which circumvented the need for estimates of absolute rates of change. This latter study (Moran et al. 1995) was similar in design to the Hafner et al. (1994) study of rates of molecular change in gophers and lice, and the results of both studies showed that rate of evolution in the parasite was faster than that of the host. I investigated the phylogenetic relationships of several lineages of endosymbiotic bacteria associated with the gopher-louse system. To do this, I sequenced the 16S rRNA gene (1,531 bp) for clones extracted from chewing 77 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. louse samples. The bacteria studied were closely related to the following genera: Acinetobacter, Facklamia, Legionella, Pseudomonas, Staphylococcus, and Xanthomonas. The gamma subclass of the Proteobacteria (which contains Acinetobacter, Facklamia, Legionella, Pseudomonas, an d Xanthomonas) has been implicated in numerous endosymbiotic relationships with insects. Members of the genus Staphylococcus are n o t known to be endosymbiotic, but were prevalent in most chewing louse and louse egg extracts, and thus warrant further investigation. DNA sequences for these taxa were used to generate phylogenetic hypotheses that could be compared to well established host phylogenies. This is the first step in determining whether any of these lineages have shared a long history of association with chewing lice or pocket gophers. MATERIALS AND METHODS The specimens collected and methods of DNA extraction, PCR, cloning, and sequencing, are provided in chapter 3. Contiguous DNA fragments were assembled with Sequencher 3.1 and consensus sequences were generated for each clone. Clones of 16S rRNA sequence were aligned with the computer program ClustalX (Thompson et al. 1997) to determine the optimal alignment. ClustalX is a global alignment program that aligns similar sequences first, based on a guide tree, before aligning more divergent ones. DNA sequences for outgroup taxa were downloaded from the National Center for Biotechnology Information (GenBank). Sequences identified as Staphylococcus were analyzed separately from those identified 78 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. as gamma-Proteobacteria. Complete sequences are given in Appendices II and HI. Phylogenetic Inference I used the best-fit maximum-likelihood (ML) method described by Cunningham et al. (1998) to choose the most appropriate ML model of nucleotide evolution. This method incrementally increases the number of parameters in the ML model until the addition of new parameters no longer increases significantly the fit between the model and the data. I used the computer program ModelTest (Posada and Crandall 1998) as a guide to select the best-fit ML model. This program calculates ML scores with 64 nested ML models and determines the best-fit model based on hierarchical likelihood- ratio tests. The selected model and parameter estimates were then entered in PAUP* (Phylogenetic Analysis Using Parsimony, Swofford 2000). A heuristic search was performed, with random sequence addition and TBR branch swapping. The resulting tree topology was then used to re-estimate the ML parameters. Another heuristic search was performed (with 10 random sequence additions) with the new parameter estimates. If the resulting tree was different from the previous tree based on -In likelihood score then, the process was repeated until the resulting tree score did not change in successive iterations. This successive approximation generates the best estimates for the ML parameters and helps to insure that tree searching does not stop at a local optimum. I used multiple outgroups in my analyses selected from well established phylogenies of 16S rRNA 79 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. sequences (Rainey et al. 1994; Lambert et al. 1998). For comparative purposes, I used the maximum-parsimony (MP) method, as implemented in PAUP*. Bootstrap analyses (1,000 replicates) were generated to determine the relative level of support for individual nodes. Comparing Host and Parasite Phylogenies Bacterial phylogenies generated in the above analyses were compared statistically to well-established louse phylogenies generated in previous analyses (Hafner et al. 1994; Hafner and Page 1995; Page et al. 1995; Page and Hafner 1996). Component analysis (e.g., Page 1990a; 1994) treats the parasite phylogeny as a lineage, rather than as a set of codes, and maps the parasite tree onto the host tree. The major limitation of this method, as implemented in COMPONENT (Page 1993a), is its poor ability to deal with host-switching events. Recognizing this, Page extended the algorithm for reconciling trees to incorporate host switching, and he implemented the method in the computer program TreeMap (http:// taxonomy. zoology.gla.ac.uk). A major advantage of component analysis (COMPONENT and TreeMap) is that it allows statistical tests of similarity between host and parasite phylogenies that take into consideration both cladistic topology and branch lengths. For these reasons, component analysis was used to determine the amount of similarity between host and parasite phylogenies. Although significant topological concordance of host and parasite (or host and symbiont) trees is unlikely to result from chance, it is possible that 80 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. recent host switching may produce spurious congruence, especially if closely related parasites preferentially colonize hosts that are closely related. Similarly, incongruence between host and parasite phylogenies may result from differential survival of multiple parasite lineages (lineage sorting), rather than host switching (Page 1993b). I mapped the bacterial phylogeny onto the phylogeny of their chewing louse hosts with the computer program TreeMap, which reconciled the host and parasite trees by maximizing the number of cospeciation events and minimizing the number of sorting events (i.e., host-switches, duplications, and extinctions). Identification of probable host-switching events may reveal whether colonization of new hosts by bacteria is simply opportunistic (nearest neighbor), or whether these symbionts are tracking a particular resource in the host environment. RESULTS Maximum Likelihood Models The computer program ModelTest (Posada and Crandall 1998) picked the Tamura-Nei (Tamura and Nei 1993) model of nucleotide evolution for both the Staphylococcus and gamma-Proteobacteria data sets. The Tamura- Nei m odel allow s for tw o rates of transitions (A<->G an d C«-»T), one rate for tranversions, and it allows unequal base frequencies. ModelTest determined that the addition of both an invariant sites parameter and a variable sites parameter (according to a gamma distribution) significantly increased the fit of the model. An invariant sites parameter assumes that some nucleotide positions do not vary across taxa in the study. The variable sites parameter 81 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. assumes that some nucleotide sites are more variable than others, and this pattern can be estimated by a gamma distribution. The shape of the gamma distribution is described by a shape parameter, a. W hen alpha is small (e.g., a = 0.05) most sites are evolving slowly but a few sites are changing rapidly. As alpha increases (e.g., a = 10) the distribution becomes centered around a value of 1.0 and all sites are evolving at nearly the same rate. Lower alpha values are most common (a = 0.03 to 0.07) in DNA sequence data. ModelTest does not perform likelihood ratio tests between all pairwise comparisons of the 64 nested models. Rather it uses a flow diagram that begins with the most specific model (Jukes-Cantor) and tests the addition of new parameters. For instance, if the addition of unequal base frequencies to a Jukes-Cantor model is found to produce a significant difference in negative log likelihood (-InL) score, the new model (in this case an F-81 model; Felsenstein 1981) is chosen over the null model (Jukes- Cantor). Sometimes the structure of the tests used in ModelTest can result in over-parameterized estimates of the best-fit ML model. To reduce the risk of choosing a model that is too parameter-rich, I tested the model chosen by ModelTest (Tamura-Nei plus invariant sites plus gamma, or TrN+I+G) against many permutations of simpler models. The removal of any parameter associated with the TrN+I+G model produced significantly worse -InL scores based on likelihood ratio tests (p < 0.05), thus showing support for the best-fit model selected by ModelTest. 82 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Gamma-Proteobacteria Analysis The parameter estimates provided by ModelTest were used in the first iteration of ML analyses in PAUP*. A heuristic search provided a single ML tree estimate upon which new ML parameters were estimated. These varied slightly from those generated previously by ModelTest and were used to perform another heuristic search (10 replicates). In this analysis, the resulting tree topologies varied among the ten replicates and the resulting parameter estimates varied again from the previous estimates. Because it is advantageous to re-estimate the parameters on increasingly better trees until the estimates stabilize, the parameters were estimated again and another heuristic search was performed (10 replicates). The resulting parameter estimates did not change during this iteration and the 10 heuristic search replicates converged on the same tree topology. The best ML tree (Fig. 5.1) showed that the clones extracted from chewing louse samples were nested within the gamma-Proteobacteria sequences downloaded from GenBank. The uncorrected sequence divergence ranged from 0.0 to 16.8 percent (Table 5.1). Parsimony analysis (equally weighted) produced an almost identical topology (redundant taxa were primed from this analysis). Bootstrap support for clades was either very low (shown as polytomies) or moderate (Figure 5.2). Figure 5.2 shows the 50% majority rule consensus tree of all bootstrap replicates (n = 1,000) and polytomies represent clades found in fewer than 50% of bootstrap replicates. 83 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. —10 Changes Acinetobacter sp.l X 8 1 6 5 9 Acinetobacter johnsonii X 8 1 6 6 3 f ex. Geomydoecns expansus 8 8 -6 r Acinetobacter calcoaceticus X 8 1 6 6 1 I Acinetobacter calcoaceticus X 8 1 6 5 7 — Acinetobacter Iwoffii X 8 1 6 6 5 — ex. Geomydoecus geomydis 9 1 -3 1 Acinetobacter haemolyticus X 8 1 6 6 2 ex. Geomydoecus geomydis 9 1 -1 1 - ex. Geomydoecus geomydis 349 -1 9 • ex. Geomydoecus geomydis 3 4 9 -1 5 ex. Geomydoecus expansus 8 8 -1 7 ex. Geomydoecus geomydis 9 1 -1 8 ex. Geomydoecus geomydis 9 1 -2 8 Acinetobacter baumannii X 8 1 6 6 7 HAcinetobacter baumannii X 8 1 6 6 0 r Acinetobacter junii X 8 1 6 6 4 *■ Acinetobacter junii X 8 1 6 5 8 * 'inetobacter radioresistens X 8 1 6 6 6 Acinetobacter calcoaceticus X 8 1 6 6 8 ------ex. Geomydoecus expansus 8 8 -5 -2 ex. Geomydoecus geomydis B -22 ------ex. Geomydoecus expansus 88 -5 -1 ex. Geomydoecus geomydis B-26 — ex. Geomydoecus geomydis 9 1 -1 9 ■ ex. Geomydoecus geomydis 349-11 Pseudomonas aeruginosa A F 2 3 7 6 7 8 Figure 5.1. Maximum likelihood phylogeny for gamma Proteobacteria using the Tamura-Nei + I + G model (see text for details of m odel). 84 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 0.0432 0.1584 0.0426 0.0404 0.1850 0.1106 0.1147 0.0439 0.0371 (table cont.) 0.0282 0.12350.1563 0.1222 0.1127 0.1174 0.0302 0.0452 7 8 0.0303 0.0468 0.1562 0.0274 0.0274 0.0247 0.0247 0.1764 0.1764 0.0302 0.0282 0.0282 0.0433 0.1503 0.1503 0.1497 0.0254 0.0254 0.0200 0.0282 0.0137 0.0220 0.0193 0.0303 0.0193 0.03630.03910.0199 0.0282 0.0488 0.0248 0.0014 0.0445 0.0350 0.0445 0.0350 0.0432 0.0363 0.0282 0.0323 0.0062 0.0275 0.0275 0.0467 0.0227 0.0193 0.0275 0.0275 0.0426 0.1167 0.1281 0.1235 0.1235 0.1145 0.0363 0.0391 0.0199 0.0323 0.0227 0.0254 0.0254 0.1525 0.1525 • 0.1508 0.03770.1249 0.0377 0.1249 0.0227 0.1293 0.1235 0.0363 0.0234 0.0234 0.1167 0.0364 0.0351 0.0234 0.0234 0.0316 0.0000 0.0371 0.1494 0.1115 0.1154 0.1154 0.1184 0.1127 0.11900.0414 0.1201 0.0324 0.1201 0.0324 0.1232 0.1174 1 2 3 4 5 6 0.0357 0.0364 0.0275 0.0303 0.03030.1536 0.0337 0.1470 0.0220 0.1543 0.0220 0.1543 0.0186 0.1540 0.0268 0.0268 0.0419 0.03640.1360 0.0302 0.1265 0.0501 0.0323 0.1237 0.0296 0.0386 0.0323 0.1519 0.0296 0.0344 0.1299 X81666 0.0426 X81662 0.0323 0.0371 0.0281 0.0281 X81657 0.0412 X81668 X81661 0.0412 0.0364 X81667 0.0357 0.0316 X81660 AF237678 0.1320 0.1135 349-15 91-31 0.0379 91-28 0.0317 0.0345 91-11 0.0323 0.0379 0.0289 0.0289 0.0151 88-5-1 0.1785 0.1765 0.1775 0.1775 0.1775 X81663 0.0289 0.0358 0.0351 0.0351 X81665 X81658 0.0391 0.0289 0.0226 0.0226 0.0268 0.0336 0.0336 X81664 X81659 Table 5.1. Uncorrected pairwise distances (P-distances) for 27 taxa of the gamma subclass of the Proteobacteria. Geomydoecus geomydis-349-19 Geomydoecus geomydis-91-\8 Geomydoecus expansus-88-5-2 Geomydoecus geomydis-91-19 Acinetobacter calcoaceticus Acinetobacter johnsonii Acinetobactersp.l Acinetobacter radioresistens Acinetobacter baumannii Acinetobacter baumannii Geomydoecus geomydis-B-26 Geomydoecus geomydis- Geomydoecus expansus-88-6 Geomydoecus expansus- Geomydoecus expansus-88-17 Geomydoecus geomydis- Geomydoecus geomydis- Geomydoecus geomydis-349-11 Pseudomonas aeruginosa Geomydoecus geomydis-B-22 Acinetobacter junii Acinetobacter calcoaceticus Acinetobacter junii Acinetobacter haemolyticus Geomydoecus geomydis- Acinetobacter Iwoffii Acinetobacter calcoaceticus Bacterial Taxa 11 Numbers following taxon names refer to National Center for Biotechnology1 Information (NCBI) accession numbers 14 15 12 13 10 or clones numbers. 2 3 19 18 16 17 8 5 4 21 23 24 7 9 6 20 22 25 26 27 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. o CM CM CO 0 0 CO CO o s c m i n CO p N NO TP ON ON r H o Cm o 1-H r-H r H o 1-H o o o i-h o 1-H 1-H o o u o © O © © d © d d O) 3 CM CO 00 VO in o vO Cm o Cm in in CO Tp NO CM O in i n t P t p Tp m tP CO CO in Tp t p r-H iH i-h i-h i-h 1“H i-h rH P-H i-h O d d © O d d d o d CM ao in o m vO CM CO Cm NO Cm 0 0 Cm ON 00 Cm O n in CM sO o in t p tP t p oo TP CO t p in tP in 1-H i- h r-H ^ h r-H ^ h p-h i- h i- h © d o d d d O d d d d CM 0 0 O s t P CO in o 0 0 Os oo in CM GO CO CO TP ^H CMCM O n t p p H oo Cm Tp Tp o o Cm o O y—* o 1“H i - h ^ H o © o i- h o t—« r—• o o 1-H O* © o ’ d o ’ o ’ o ’ o ’ o ’ o ’ d d NO CM CM CO Cm ON 00 o 0 0 ON in VO CM 0 0 in CO in CO CO CM ON t p CO 0 0 oo o Tp Tp o r-H o Cm o o f-H f“H o p H o 1-H ^ H o o o r-H o f-H f-H o o 1-H d © d d © o ’ d o d d d d o ’ CM Cm 0 0 1-H ^ H Cm m in CM Cm fH CM Tp ON o o p H CO CM CM 1“H tp O 0 0 CO NO 1-H 1“H CM CM in Tp CM CM CM Cm CM CO f-H CM CM TP f-H 1“H 1-H f-H 1-H f-H i- h 1-H r-H o o f-H 1-H o o ' o ’ o ’ d o ’ o o ’ d o ’ o ’ d o ’ d o ’ Tp vO ON vO ON vO 0 0 vO 1-H Os Tp CO NOCm T p 0 0 O n 0 0 CO ON r-H VO ON o 00 in CM Os Tp TP CM CM CM m Tp CO CM CM 00 CM 1“H CM CM CM CM r-H o o f-H 1-H o o O 1“H © PH f-H © O p H © o o © o* o ’ o ' © o ' o ’ o ’ o © o ’ o ’ o CO in CM Cm CM in O n O n CM 00 in NO 1-H o oo m Tp ON 0 0 in CO 1-H o 00 NO 00 CO ON in CO CM CO CM CM CM in in CO CO CM CM CM 1-H 1-H CO CO CM o 1-H O O 1-H 1-H o O O 1-H O 1-H 1-H o o d d O o ' o ’ o ' d O O d O d O d o ’ O sO CO Cm NO NO Cm 0 0 NO O s o c o ON i n vO VO t p o o 1-H c o CM Cm Cm CO ON ON OO ON O n vO Cm 1-H Cm 1-H i n CO CM CM O O in Tp o o o Cm o 1-H CM o o CM O o r-H O O 1-H 1-H o o o t h p 1-H 1-H O p © © O d © d © © © d o d d o d d d in 1-H o r-H in r-H Cm o in CM 0 0 GO GO CO TP 1-H NO NO ON TP CO in Cm NO in TP ON OO NO NO NO TP o CO ON CO CM O CO CM CM CM in in CM CM CM N CM CM CO CM CM OO p 1-H O O 1“H 1-H o O o f H o 1-H 1-H o O 1-H o ’ o ’ d d o o d d d d d d d d d d o d onNn^in«Noo»OriNn«inifits INn«lD'eNe09iHrlpHplrlrtpCrlrlnWNPIN«NNN 86 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Ntv ON vO »-H N CN CN lft ON N O CNi-* OO NO CO o m «—• CN i-h CO N i-h i-h O O 0*0 CN ON O CO 00 in i-H i - h co CN i h 1/1 N O 1-H T~« o o o o o CN c o T}1 OO o ON N* CN t>* oo CN o t*H 1-H O ^ H CN I-H 1-H o o I-H o ’ o* 0 * 0 * o CO CO CO on 0 0 CO CO ON oo NO co oo t v nO NO N N VO 1-H I-H 1-H 1-H I-H H o o o o o o ON CO CN t v 00 oo o CO CO r-H m c o o n oo t v o |H 1-H 1-H o f-H f-H o 1-H 1-H o o ^H o o d o 0 * 0 © CN y—, in O n i-h CN NO NO ON in CN NO in On CN FHNO ^H1-H FHi-h O CN o FHo r*H FH OO r-H o o o* © o ’ 0 * 0 o ’ OfHC40^tAvONC09\OiNCNO«U)vOts • NO^'lftvO^oOGvr'HpHiNHrtfN^iNr^NW NNNM W N 87 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Acinetobacter sp.l X 8 1 6 5 9 82 Acinetobacter johnsonii X 8 1 6 6 3 ex. Geomydoecus expansus 8 8 -6 Acinetobacter calcoaceticus X 8 1 6 5 7 64 Acinetobacter Iwoffii X 8 1 6 6 5 57 Acinetobacter haemolyticus X 8 1 6 6 2 ex. Geomydoecus geomydis 3 4 9 -1 9 52 ex. Geomydoecus geomydis 3 4 9 -1 5 ex. Geomydoecus geomydis 9 1 -2 8 67 ex. Geomydoecus expansus 8 8 -1 7 ex. Geomydoecus geomydis 9 1 -3 1 ex. Geomydoecus geomydis 9 1 -1 1 100 51 Acinetobacter baumannii X 8 1 6 6 0 Acinetobacter junii X 8 1 6 5 8 100 Acinetobacter radioresistens X 8 1 6 6 6 Acinetobacter calcoaceticus X 8 1 6 6 8 ex. Geomydoecus expansus 8 8 -5 -2 65 65 90 ex. Geomydoecus geomydis B -2 2 72 ex. Geomydoecus expansus 8 8 -5 -1 ex. Geomydoecus geomydis B -2 6 Pseudomonas aeruginosa A F 2 3 7 6 7 8 ex. Geomydoecus geomydis 3 4 9 -1 1 ex. Geomydoecus geomydis 9 1 -1 9 Figure 5.2. Fifty percent majority rule consensus tree of 1000 bootstrap replicates using the parsimony optimality criterion for species of gamma Proteobacteria. Numbers at nodes represent bootstrap support. 88 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Staphylococcus Analysis The parameter estimates provided by ModelTest were used in the first iteration of ML analyses in PAUP*. A heuristic search provided a single ML tree estimate upon which new ML model parameters were estimated. These varied only slightly from those generated previously by ModelTest. They were used to perform another heuristic search (10 replicates) which provided the same topology in each of the ten replicates. The resulting parameter estimates did not deviate from the previous estimates and therefore the process of iteration was terminated. The best ML tree (Fig. 5.3) showed that the clones extracted from chewing louse samples were nested both within and outside the Staphylococcus sequences downloaded from GenBank. The uncorrected sequence divergence ranged from 0.0 to 12.4 percent (Table 5.2). Parsimony analysis generated a similar phylogenetic tree that was well supported. Figure 5.4 is the 50% majority rule bootstrap tree that was generated in the parsimony analysis, and again, bootstrap support was either very low (values less than 50% are shown as polytomies) or moderate based on 1,000 replicates (Figure 5.4). Some of the clones in this analysis were closely related to known taxa whereas others seem to be phylogenetically distinct groups. Several clones isolated from chewing lice found on the pocket gopher Thomomys bottae are basal to the known Staphylococcus species shown in Figures 5.3 and 5.4. When 16S rRNA sequences from these clones were queried against the NCBI database using BLAST searches, the most similar taxa were those used in this analysis. In 89 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. — 5 Changes ex. Thomomydoecus minor -21 ex. Thomomydoecus m inor -19 ex. Thomomydoecus minor -2 7 ex. Thomomydoecus minor -2 5 ex. Thomomydoecus minor -1 7 _ ex. Thomomydoecus minor -3 0 - e x . Geomydoecus setzeri-7 - ex. Geomydoecus geomydis-6 - e x . Thomomydoecus minor -2 0 e x . Geomydoecus centralis-11 . ex. Geomydoecus centralis-1 ex. Thomomys bottae*-2 e x . Thomomydoecus minor -8 e x . Thomomys bottae*-7 ex. Thomomys bottaeM e x . Geomydoecus expansis-21 ex. Geomydoecus geomydis-17 ex. Thomomys bottae*-5 I e x . T hom om ys bottae*-10 ex. Geomydoecus expansis-7 Staphylococcus succinus A F 0 0 4 2 2 0 e x . Geomydoecus expansis-9 ex. Geomydoecus expansis-1 r e x . Geomydoecus expattsis-10 - ex. Geomydoecus expansis-5 - e x . Geomydoecus expansis-8 Staphylococcus gallinarum D 8 3 3 6 6 — Staphylococcus saprophyticus L 2 0 2 5 0 e x . Geomydoecus panamensis-9 Staphylococcus cohnii A B 0 0 9 9 3 6 Staphylococcus pasteuri A B 0 0 9 9 4 4 CStaphylococcus wameri A J2 7 6 8 1 0 ex. Geomydoecus expansis-21 ex. Geomydoecus setzeri-3 £ex. Geomydoecus expansis-25 — Staphylococcus aureus X 6 8 4 1 7 1- Staphylococcus sciuri A F 0 4 1 3 5 8 Staphylococcus pulvereri A B 0 0 9 9 4 2 1*1 Staphylococcus vitulinus A B 0 0 9 9 4 6 •— Staphylococcus lentus D 8 3 3 7 0 Bacillus subtilis A J2 7 6 3 5 1 • Macrococcus bovicus Y 1 5 7 1 4 Figure 5.3. Maximum likelihood phylogeny for Staphylococcus species using the Tamura-Nei + I + G model (see text for details of model). Clones marked with an asterisk could be from either of two hosts (Thomomydoecus minor o r Geomydoecus centralis). 90 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 7 0.0071 0.0045 0.0039 0.0311 0.0227 0.0229 0.1208 0.0234 0.0266 0,0045 0.0233 0.0343 0.0350 0.0188 0.0227 0.0361 0.0214 Thomomys Thomomys bottae. (table cont.) 0.0110 0.0097 0.0071 0.02750.0273 0.0263 0.0248 0.0259 0.0246 0,0220 0.0259 0.0233 0.0259 0.02530.0315 0.0251 0.0228 0.0232 0.1135 0.1126 0.0162 0.0136 0.0084 0.0045 5 6 0.0297 0.03500.0337 0.0104 0.0343 0.0091 0.0295 0.0248 0.0095 0.0318 0.0360 0.1129 0.1237 0.0317 0.02460.0080 0.0220 0.0324 0.0087 0.0312 0.0308 0.0331 0.0084 0.0304 0.0291 0.0239 (listed by louse taxon). Numbers following 0.0363 0.0356 0.0376 0.0363 0.0291 0.0337 0.0323 0.0310 0.0094 0.0304 0.0291 0.0304 0.0292 0.1005 0.0992 0.0344 0.0311 0.0298 0.0155 0.0130 Staphylococcus 0.0019 0.0026 0.0350 0.1136 0.1143 0.0357 0.0102 0.0108 0.0343 0.0350 0.0337 0.0337 0.0324 0.0337 0.0337 0.0343 0.0330 0.03300.0298 0.0318 0,0998 0.0304 0.0052 0.0032 0.0045 0.0323 0.0303 0.0310 0.0376 0.03560.0369 0.0337 0.0343 0.0330 0.0343 0.0323 0.0330 0,0317 0.0328 0.0308 0.0315 0.0302 0.0356 0.0337 0.0343 0.0330 0.0330 0.03130.0092 0.0293 0.0071 0.0300 0.0078 0.0287 0.0064 0.0298 0.03560.0356 0.0336 0.0343 0.0330 0.0356 0.03360.0389 0.0343 0.0369 0.0330 0.0339 0.0319 0.0325 0.0350 0.0330 0.03570.03300.0356 0.03370.0336 0.0310 0.0344 0.0317 0.0317 0.0331 0.03040.0362 0.0317 0.0342 0.0252 0.03220.0107 0.03220.0114 0.0302 0.0087 0.0329 0.0094 0.0308 0.0315 0.0100 0.0215 0.0195 0.0330 0.0310 0.0317 0.0304 0.0213 0.1018 0.0330 0.0311 0.0317 0.0324 0.03170.0343 0.0297 0.0323 0.0304 0.0330 0.0350 0.0330 0.0337 0.0324 0.0253 0.0356 0.0336 0.0343 0.0330 0.0259 0.0234 1 2 3 4 0.1111 0.1143 0.0304 0.0039 0.0643 0.0616 0.0350 0.0331 0.0337 0.0324 0.0078 0.01360.0597 0.0122 0.0473 0.0085 0.0609 0.0339 0.0590 0.0582 0.0629 0.0583 0.0349 0.0329 0.0100 0.0107 0.0948 0.0623 0.0603 0.0595 0.0609 9 6 0.0630 0.0363 0.0344 0.0350 17 0.0608 21 0.0595 10 24 0.0622 9 0.0642 5 0.0589 25 4 0.0595 8 0.0570 24 25 17 0.0316 20 0.0623 19 27 30 0.0329 0.0310 0.0575 8 0.0635 14 1 0.0623 7 3 0.0622 2 5 7 0.0596 4 0.0590 15 Table 5.2. Uncorrected P-distances for 42 taxa of Bacterial Taxa Thomomydoecus minor Thomomydoecus minor Thomomydoecus minor Thomomydoecus minor Geomydoecus setzeri Geomydoecus geomydis Thomomydoecus minor Thomomydoecus minor Geomydoecus centralis Geomydoecus expansus Bacillus subtilis Thomomydoecus minor Thomomydoecus minor Thomomys bottae* Geomydoecus expansus Geomydoecus geomydis Geomydoecus expansus Staphylococcus gallm rum Geomydoecus centralis Staphylococcus Jentus Geomydoecus expansus Geomydoecus setzeri Staphylococcus pasteuri Staphylococcus warneri Geomydoecus panamensis Geomydoecus expansus Staphylococcus cohnii S. saprophyticus Staphylococcus sciuri Geomydoecus expansus Thomomys bottae* Thomomys bottae* Geomydoecus expansus Geomydoecus expansus 7 Staphylococcus succinus Thomomys bottae Staphylococcus pulvereri Staphylococcus vitulinus Thomomys bottae* Geomydoecus expansus Staphylococcus aureus Macrococcus bovicus 5 7 10 15 taxon names refer to clones numbers. Asterisks1 denote clones that came from4 lice of the gopher 6 3 8 9 11 12 13 16 17 2 21 20 36 14 37 18 19 25 31 38 39 40 42 22 23 24 26 27 34 30 32 29 35 41 25 33 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. C v O O O v CO rH oo00 CN| cooo rH CM NO Tf tn CM Cm1-H 00 NO o CO CM V O C S fH CM OS 00 NO NO tM 00 On tM ON rH rH oo cm o f"H rH 00 Ov CmOn o w o o fH o rH o o o o CM CM o CM co o CO o CO |-H rH CM o l-H u p o o o o o o o o o o o o o o O o o o o o o o o i-h rH o d d o d d d d d d d d d o ’ d d d d d d d d d d d d 3 (0 NO t-« O CMCMON O n CO NO CO NO CM ON in oooo ON CO NO rH CM NO o tS*Cm ON NO Cm r r CM 1-H Cm CMin in NO Cm Cm oo CM NO CMCm CO Cm rH Cm NO o CM ON CM O CM CM COCM CM CM CM CM CM CM CM o i-h CM 1-H CO CM CO CM CO CM fH CO o rH o OOO o O o o o o o O O o o o o o O o O O O p p»-h i -h d o o o * o d o o* o* d o d d O d d d o d o ' d o’d o o ’ o ’ d oo00ON ON CmTt ON o O n CO Cm 0 0 CO rH rH tn o CO NO CO in CM Cm oo H NO CO CO 0 0 00 NO l-H CM ON |-H Cm ON CM O ON in l-H o 0 0 Cm ON rH o oo CM O CM o T f CO CO1-H 00 rH CO CM CM CM o CM ?-» O o o CM rH o o o o CM CM 1-H CM CO o CO o co rH CO o rH o o p p o o o p o o p o o o o o o O O p o O o o p o pl-H rH o’ d d o’ o’ o’ o’d d o' d o ’ d o’ d o o ’ o ’ d o d o ' o' o* o’ d o’ d d CO 1—* NO 0 0 NO O ^ in CO oooo CM 0 0 o NO CO CM CM T* 0 0 NO tn in m CO CM ON NO r H o NO Cm NO O n GO H} ^HH*CMrHrHrn CM 00Cm Onin oooorH00 oo Onoooo NOoo CM COco Cmo in CO OOOn in On00ON ^ 00CMONCONOin vo0000ONONtn TJ1o co oo 00 in CMCM o Cm rHrHo CMrHCM rH rHCMrHo o o rH rHCMCMrHCMCOrHCOrHCOCMCMCOr H rn o O o O o p o o O o p po p o o o o o o O p O o o p O O p l-H r H d d d d d o d d d d d d d d d d d d d d d d d d d d d d d d d in m Cm ON CM CM CM Cm o 0 0 ON o T}«oo m oo CO NO OO CM NO CM CO CO Cm o c o ON CO ON O rH O n o O n no O o l-H rH o oo ON o o ON NO NO O rH o CM o c o CO CM CO O n CM o rH CM O CM CM CM n f S CM CM o rH rH l-H CM CM CM CM l-H CM CO CM CO CM c o CM CM CO O rH o o O o OOOOO OO p OO o o O o o O p OO O p O o o O p 1-H l-H d d d d d d o d d d d d d d d d o* o ’ d d d d d d d d d d d d d d tn oo Cm CM CM o CM rH r ji o Cm CM Cm o Cm •**» r* Cm ON QO in CO l-H O n in o •M* o NO CO Hji Cm rH CM m CM CM Cm 00 r i NO CM tn rH l-H O n CM CM ON in OO ON CM CM CM CM rH o ON O O CM CM o CM CM CM r n CM CM CM o rH n CM CM CM CM CM CM l-H 1-H CO CM CO CM CO CM CM CO rH rH O O O O o O O OOO o O o O o O O OO o o o p pO o o o o O p 1-H l-H d o ’ d o ’ d d d o d d d d d d d d o ’ d d d d d d d d d o ’ o ’ d d d o* d OO NO 00 OOO CMin Tt< m n o CO 1-H CMCOCm oo r n O n CMin O n in NO 00 CMo tji Cm CO r H in c m in o CMc o o 1-H O CM CMin CMCOCM o o 1-H CMr n Cm OO CMl-H CO r H Ht* in c o r H O o o o CMCMo COCMCO 1-H CMCMCMo l-H r H CMCMCMCMCO CMr H CMCOCMCOCMCOCMCMCO r H CM o o o o O o o o poooOooOooOO O O o p o O o O o o O p r H r H o d d d d o d d o d d d d d d d d d d d d d d d d d d d d d d d d d O r H CM CO to NO Cm CO O n o rH CM CO in NO Cm 00 ON o rH CMCO in NO 00 On o fH CM HPin'finvoM nffirti-iON r H rH rH rH r * rH r n r n ^ rH CM CM CM CM CM CMCM CM CM CM COCOCOCOCO co COCOCO CO 92 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. c CO T p r -i ON NO TP r - »1-H T p oot v CO 0 0 CO vO 0 0 CM LO tM ooO n in TT O n o o 0 0 o o ON i- h cm Cm tM O n O CM o o oCM CM o CM CO o CO o CM i- h rH CMo i- h U o o o o o o o o o o © o o O rH i—» J£ o o o o o o © o o o © o oo ’ o o © 3 n3 ONCMONin CM oo in ONCMCOin CMo o ONo TP NO NO CMin 1-H CMNOONNOo NOo Cm 00 On in T p r H i- h r H ♦H CMCM r H CMi-H CO r H CO r n r H CMo r H © o © O o o p o p O o O p O o O r H i - h o o ’ © o o ’ o ’ o ’ © o ’ o o o ’ o ’ o* o OO o ON in CO in r H O n o T p 00 CM in O n o CM tP CM Cm VO NO r n co in VO C m OO in CM CONO o N O C m r H oo 00O n NO T p CM o l-H r r r H CMCM r H 1-H CO r H CO r H CO r n r CM o r H o o o Cr O o o O o o o p O o o OO r H t- h o*o’ © © O © o*o ’ oo ’ o ’ o o ’ o© ’ o’ O O © TpCm in 00 in r H tn ONtn CO co inO n o o o Cm vO O n o CM o On Cm 00 O' o 00 m in 00 CMONCMON TpCO O CMONCO r H O r H r n r r CMCMCMr H 1—H CO 1-H CO r H CO CMCMCOo r n CM OO OO o o o o O o OO O O p o O o 1-H © © o* © © © o ’ o ’ o o ’ O* © o o’o’o’o’©o © r H r H tn CO vO o inCMCm ooCMNO00 oo00 ooVOCMCO o O 00 CMON r n o o 0O CMO CMr p o H p COCO 1-H CO COCO CM r H r n o o o CMCM CMCO o coo co1-H 1-H CO o 1-H O o O o o p o o O p o p o op o o pr H rH o o O o o o o o ’ o*o ’ O o o o o ’ o o ’o o’ oo ’ 00 TP HpCOCO in CMoo00 CO r H CMVOCMCMCMco1-H o o Hp in CO 1-H o CMHP tn tnr n Cm TpVOCMin 00 CMCm in CMTP NO1-H O o CM CMCMr H o o oCO CM1-H CMCO o CO r H CO 1-H CMCO o CM CM o p Oo o o o o O Oo O o p o o O op o1—1I— o’ o ’ o o’o o o o o ’ o’ oo ’ o* o o o ’ oo ’ o* o oo ONCm tp m oo ONTp CMoo CO r H CM1-H oo Tp00 in coCm TpON T-H CMO n O n CmvO oo ON ON C m m o CMCMO n CMr H r H CMO n O n 00 ON r H o t-H r H r n o o oCMCM CMCO O coo CO 1-H CMo 1-H O O p p oo p O o O op O o O o op p oo 1-H o o o oo o o 0 o O o o o o o o o o o o o o o CM vO CO O n O n COo o Cm COvO TP o Hpr n Cm 00 oo oo in ONCM00CO O n COo tP CMCMO n O n O n o TP O n o tM o ONo CO O n o tn VOm CO O i- h r H r H r H r H o O c? r H CMr H o COo COo CM1-H r H CMo r n OO O o o o o O p OO o o p o p o p o o o O 1*H 1-H o o ’ o ’ o ’ o o o o* o o o o* o* o o o o ’ o ’ o* o ’ o o o O CM CMONO n CO VOCOvO CMO n 1-H m 0O coONCOVO r H CMvO O TP Cm ON ^P Cm O n Cm Cm m Tpin vO Cm Cm ONCm TpvO CMCm COCm r H tM vOO CMON CM CM CMCMCM CMCMCMCMCMCMO r H CMr H COCMCOCMCOCMl-H COO OOO o OO o oO p o o o O O O p O o O O p 1-H r n o o o o’o ’ o o o’ o O o ’ o o oO ’ o ’ o o ’ o ’ o o’o o*o OHNO^I/1\ON«OO n O CM CO m vO C>H0 0 O n o rH CMCO in VO tv COOn o rH CM nfMO^lfivONOOO'pHriiHP^HfHHfHrljH^HN CMeM CMCMCMCMCMCMCMCOCO CO COCO CO COCOCOCOTP TP TP 93 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. O n o o cm ON CM ON Cm c NC t p O n ON ^ p NO in CO o CO CM o N ON o co © © o © o © o u o o o o ’ © © o .2 X>cc , ■CM ON in CM Tt* 00 CO «—«N o CM N Tp CO o CO 1—< CO o co o o p o o o o o ' o o ' o o © O Cm CM 1—* o CM in CM CO CO o NO t t ON o c TT CM co CO o CO o CO CM o rM O n co o o o o o o o © O o o o ' o o o ' o ’ o* o Tp o 1—■* Cm Tp ON Tp 00 CM CM ON O n in NO o CM CM CM CM CM CM CM o CM 00 O CO o o © o O O o o o *—• o ' o ' o o o o ' o ' o ’ o ’ o ’ TpT- O 00 oo oo O n CO CM TpNO Cm o o o co ON O NO NO o in CO <—* CO o CM FCM oo ON O o p o o o o O o O o ' o 'o ’ o ’ o ' o o o o ’ o ' © t—»00 00 o in Tp o o in CO «—> o VO 0O CM ON in oo CO vO in vO o CM CM CM CM CM CM CM T“^ CM o CO O o o o p o o o O o *—* *—• o © o ’ o ' © © o ' © o o o © in in on CM00 oo CMCm f—1COCM 00 in Cm CO CMTT Cm 1“^ OOvO COON ON f—HCM1—4CO CMCO CMCOCM COo 1—* CM o O © o o O O p O O o «—* o o o o o ©o'o' o o o o o ’ in Cm CM 0 0 CM vO 0 0 0 0 0 0 0 0 vO CM CO o 0 0 CM o CM Tp o TT00 CO r—< 0 0 CO 00 oo CM CM 1-^ CM CO O CO o CO f—* 1-^ CO o CM o O p O o p p o o o o o o ' o ’ o o ’ o* o o o ' o ' o ' © o © o VO ON o ,,in o CM , CM CM ON in Cm TP O CM CM o CM co o TP 00 CO Cm o CM Cm cm O CM CM CM co o CO o CO T-*!—• co O r-^ CM pp oooo oooo OO o r-< *—« o ' o o o o ’ o o o ' © o o ' o o o ’ o vO CO CM Cm TP ''p 00 VO in ' p in in CO Cm ON vO O ^ Cm CM ON 1-^ CM O CO Cm CM o Cm O n vO o O O CM CM o CM CO O CO O CO CM o CM OOOO p O O O p o o o o O o o o O o O o o o © © o o ' © o ’ o CM a o ON o CM CO Tp in Cm 00 On o r>HCM CM CM CM CO CO CO CO CO CO CO CO CO CO *p Tp ^P 94 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ON CO 00 CO CM CO Cs* CO CO p © t n i n O T— ©CO p 1— o © 0.0183 0.0284 0.0826 0.1021 0.0285 0.0071 0.0743 0.0193 0.0977 0.0277 0.1057 0.1193 0.0263 0.0964 0.1182 0.0489 0.0775 0.0689 r^rlCO^invOtNOOOM-^r^t-ti^^r^rHTHr^rNCMCMCMCMCNieMrJeMCMC'leOCOCOeoeOCOCO^^^^'^^ 95 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. - ex. Thomomydoecus minor -24 - ex. Thomomydoecus minor -19 - ex. Thomomydoecus minor -27 - ex. Thomomydoecus minor -25 - ex. Thomomydoecus minor -17 - ex. Thomomydoecus minor -30 - ex. Geomydoecus setzeri-7 66 - ex. Geomydoecus geontydis-6 100 - ex. Thomomydoecus minor -20 90 - ex. Geomydoecus centralis-14, - ex. Geomydoecus centralis-1 90 - ex. Thomomys bottae*-2 - ex. Thomomydoecus minor -8 67 - ex. Thomomys bottae*-7 - ex. Thomomys bottae*-4 - ex. Geomydoecus expansis-21 76 - ex. Geomydoecus geomydis-17 64 - ex. Thomomys bottae*-5 63 81 98 - ex. Thomomys bottae*-10 ■ ex. Geomydoecus expansis-7 • Staphylococcus succinus AF004220 54 ■ ex. Geomydoecus expansis-9 ■ ex. Geomydoecus expansis-4 ■ ex. Geomydoecus expansis-10 93 ■ ex. Geomydoecus expansis-5 80 ■ ex. Geomydoecus expansis-8 99 ■ Staphylococcus gallinarum D83366 92 • ex. Geomydoecus panamensis-9 ■ Staphylococcus cohnii AB009936 ■ Staphylococcus saprophyticus L20250 60 • ex. Geomydoecus expansis-24 100 ■ ex. Geomydoecus setzeri-3 76 ■ ex. Geomydoecus expansis-25 • Staphylococcus aureus X68417 85 • Staphylococcus pasteuri AB009944 Staphylococcus wameri AJ276810 99 Staphylococcus pulvereri AB009942 61 Staphylococcus vitulinus AB009946 Staphylococcus lentus D83370 Staphylococcus sciuri AF041358 Bacillus subtilis AJ276351 Macrococcus bovicus Y15714 Figure 5.4. Fifty percent majority rule consensus tree of 1000 bootstrap replicates using the parsimony optimality criterion for Staphylococcus species. Clones marked with an asterisk could be from either of two hosts (Thomomydoecus minor or Geomydoecus centralis). 96 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. other words, these clones have 16S rRNA sequences that represent lineages distinct from those surveyed elsewhere and deposited in the NCBI database (GenBank). DISCUSSION Culture Independent Methods The effectiveness of 16S rRNA genes for detecting phylogenetic signal among Eubacteria has been debated widely. Since culture-independent sampling methods came into practice, (Hugenholtz and Pace 1996, Hugenholtz et al. 1998, and Pace 1997), the number of ribosomal DNA sequences in nucleotide databases has grown rapidly. However, some researchers have cast doubt on the effectiveness of the 16S rRNA gene in phylogenetic studies of closely related organisms. David Ward and his colleagues (Ward et al. 1998) showed that the 16S rRNA gene was insufficient to resolve the phylogenetic relationships of distinct ecotypes of bacteria associated with microbial mats communities. Furthermore, the ecologically and genetically distinct populations studied by Ward and his colleagues (differentiated based on internal transcribed spacer DNA sequences) had identical 16S rRNA sequences over their entire length (>1,500 bp). However, studies such as those by Paul Baumann and his colleagues (e.g., Baumann et al. 1997) found sufficient variation in the 16S rRNA gene to resolve the phylogenetic relationships of apparently closely related species. The discrepancy may be in the designation of bacterial "species." 97 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Because there are so many 16S rRNA sequences for Eubacteria in genetic databases, researchers using culture-independent analyses are constrained to use ribosomal sequences to identify unknown samples. As microbiologists add other genes to these databases, there will be decreased dependence on ribosomal sequences. Regardless of their usefulness in phylogenetics, ribosomal DNA sequences provide an effective "name tag" for unidentified bacteria. In my work, the 16S rRNA gene showed that the pocket gopher-chewing louse assemblage harbored many types of bacteria (see chapter 3). More specific analyses (e.g. the analysis of gamma- Proteobacteria and Staphylococcus) showed that DNA samples extracted from chewing louse samples can also contain numerous closely related taxa. Bacteria sampled from lice cluster within and among well known species of bacteria (Figs. 5.1 - 5.4). However, many of these species were described based on the amount of 16S rRNA sequence similarity between unknown samples and previously sequenced species. David Ward (1998) has discussed the flaws in this phenetic classification procedure and has encouraged microbiologists to use a more natural species concept. Therefore, the delineation of "species" for the taxa of bacteria associated with chewing lice (e.g., Acinetobacter or Staphylococcus) is problematic due to lack of information on the natural history of these organisms. Analyses of the gamma-Proteobacteria Bacterial 16S rRNA sequences that were tentatively identified as gamma-Proteobacteria (using BLAST searches of GenBank) were completely 98 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. sequenced and analyzed cladistically. Known gamma-Proteobacteria were analyzed along with the clones sampled from chewing lice. Some clones grouped within the dade of Acinetobacter species downloaded from GenBank (Figure 5.1). Three clones (88-5-1, 88-5-2, and B-22) were closely related to the Acinetobacter spedes, but were found to be sister to them rather than nested within Acinetobacter. W hen the com plete 16S rR N A sequences for these clones were queried against GenBank, the most closely related taxa were the Acinetobacter species used in this analysis rather than a group nested between Acinetobacter and the outgroup taxa. These clones could represent either a new lineage of Acinetobacter or a distinct undescribed lineage. Three additional clones appear outside the Acinetobacter dade (B-26, 91-19, and 349-11; Fig. 5.1). BLAST searches of GenBank determined that the dosest taxa to these three outliers are members of the genus Pseudomonas. Previously, these clones had been assigned to the genera Legionella, Pseudomonas, and Xanthomonas based on partial 16S rRNA sequences. Figure 5.1 was rooted with the only putative outgroup taxon known at the time {Pseudomonas aeruginosa). Figure 5.2 (analyzed after Figure 5.1) used the three outliers and P. aeruginosa as outgroup taxa. It is evident that the relationship among the four individuals has changed in Fig. 5.2 and warrants further investigation. To understand the phylogenetic relationship and taxonomic placement of these taxa, better sampling and a new series of outgroup taxa will be required. 99 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Bootstrap analysis (Figure 5.2) shows some ambiguity in the relationships of the known species of gamma-Proteobacteria. It appears that the low bootstrap support for some clades is a product of the nucleotide substitution patterns inherent in the ribosomal DNA sequences of these taxa. There are long regions of invariable sites punctuated by regions of high variability (Appendix II). Where nucleotide changes occur between taxa, the changes contain homoplasious information and provide limited phylogenetic signal. As a result, some clades remain unresolved. A genetic marker that evolves at a rate faster than the 16S rRNA gene may help to clarify these relationships. Analyses of Staphylococcus Clones that were similar to Staphylococcus based on preliminary BLAST searches of GenBank were completely sequenced and analyzed by maximum likelihood and parsimony analyses. Eleven species of Staphylococcus were downloaded from GenBank along with two outgroup taxa, Bacillus subtilis and Macrococcus bovicus. Many of the clones extracted from chewing lice clustered within the known Staphylococcus taxa (Fig. 5.3). Whereas other clones (Thomomydoecus minor 24, 19, 27, 25, and 17) appear as basal lineages relative to the known Staphylococcus species (Fig. 5.3). Still other clones (those from Thomomydoecus minor, Geomydoecus setzeri, G. geomydis, G. centralis, and those from the gopher taxon Thomomys bottae) form a unique clade nested within the known Staphylococcus species (Fig. 5.3). The most basal lineages of clones (Fig. 5.3) could be unique to the 100 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Thomomys bottae assemblage or an artifact of incomplete taxon sampling. The complete 16S rRNA sequence for every clone in this analysis was queried against GenBank and the closest matches were those found in the main clade of Staphylococcus species. These clones appear to form two basal lineages (all from Thomomydoecus minor) and one derived lineage (those from Thomomydoecus minor, Geomydoecus setzeri, G. geomydis, G. centralis, and those from the gopher taxon Thomomys bottae) related to Staphylococcus. For the same reasons stated previously, it would be premature to formally name these taxa as distinct species based solely on their 16S rRNA sequences. However, both of the basal lineages and the derived lineage appear to be quite distinct from the taxa of Staphylococcus for which 16S rRNA sequences are known (Fig. 5.4). Bootstrap analysis, based on 1,000 replicates using the parsimony optimality criterion, shows weak to moderate support for most of the clades in the analysis of Staphylococcus and Staphylococcus-like clones . It seems that the 16S rRNA gene is able to resolve interspecific relationships in this clade, but it fails to resolve relationships among more closely related taxa. A genetic marker that evolves at a rate faster than the 16S rRNA gene would be beneficial in future analyses of these bacteria. Cophylogeny Of the two phylogenetic analyses presented in this chapter, only the one involving species related to Staphylococcus (Fig. 5.3) could be examined for patterns of cophylogeny. The clones sampled from chewing lice in the 101 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. analysis of the gamma-Proteobacteria (Fig. 5.1) came from only two hosts and studies of cophylogeny must contain at least three hosts as there is only one phylogenetic relationship possible for two taxa. Patterns of cophylogeny often are cryptic to the investigator, therefore, statistical tests of similarity between host and parasite phylogenies are the only objective way to discern these patterns. In the case of the Staphylococcus species, there are five chewing louse hosts {Thomomydoecus minor, Geomydoecus centralis, G. expansus, G. p anam ensis, and G. setzeri) and 29 bacterial clones sampled from the five hosts. Preliminary analyses of chewing louse and Staphylococcus tree topologies (using TreeMap) suggest that there are only three cospeciation events and many host switching events that encompass the entire tree topology depicted in Figures 5.3 and 5.4. However, patterns of cophylogeny may exist in localized areas of the phylogeny. For example, certain subclades within the Staphylococcus assemblage may show cophylogeny with their insect hosts even though the overall assemblage does not. The difference, of course, is the scale at which one compares host- parasite pairs. Despite little evidence of cospeciation between Staphylococcus and their chewing louse hosts in this analysis, there could be evidence of cospeciation between one or two clades of Staphylococcus and their hosts. Currently, investigation at lower phylogenetic levels is hampered by incomplete taxon sampling. Future studies should sample specific Staphylococcus clades intensively by designing PCR primers that 102 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. preferentially anneal to only a few Staphylococcus species, which make it easier to amplify the target sequences. Chewing Louse and Bacteria Associations Because this study was the first to document the association of bacteria with chewing lice of the family Trichodectidae, it is not surprising that this analysis generated more questions than it answered. I have now documented that multiple taxa of both Acinetobacter and Staphylococcus are associated with the pocket gopher-chewing louse system. Whether these taxa represent multiple species of bacteria remains to be tested. At present, I cannot say whether either lineage is truly endosymbiotic. To test the hypothesis of endosymbiosis, future research should use microscopy techniques in combination with in situ hybridization, which causes target species of bacteria to fluoresce. This would allow the researcher to target certain species of bacteria to determine if they are present within the chewing louse host. It is possible that the bacteria surveyed in this study are only transient associates of chewing lice. Despite efforts to sterilize the surface of chewing lice prior to DNA extraction, it is conceivable that the bacteria described herein are actually associated with the pocket gopher or the underground burrow system in which the assemblage lives. Undoubtedly, we are only scratching the surface of a complex interaction between a mammal, an insect, and numerous microorganisms. This particular host-parasite system (gophers, lice, and Eubacteria) is exceptionally promising because of the high level of population isolation 103 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. documented for the mammal and insect lineages. Future work must include base-level characterization of the microbial community associated with pocket gophers, chewing lice, and their subterranean burrow systems. In addition, knowledge of the basic natural history of the bacteria associated with the gopher-louse system is important so that future studies of cospeciation can focus on lineages of bacteria that have biologically meaningful interactions with the (gopher-louse) system, rather than transient bacteria or forms introduced by contamination. 104 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. SUMMARY AND CONCLUSIONS This dissertation examined the possible mechanisms of cospeciation in a host-parasite system and surveyed new associates in a model system for the study of cophylogeny. Pocket gophers and their ectoparasitic chewing lice have shared a long history of association. Because of low host and parasite vagility, lineages of chewing lice have been stranded on lineages of pocket gophers for millennia. Through time, these lineages have speciated in parallel. This pattern of cophylogeny has likely been reinforced by many mechanisms, including the evolution of an intricate fit between the louse rostral groove and the hair shaft of its natural host (chapter 1). Whereas the fit between louse rostral grooves and gopher hair shafts may reinforce patterns of cophylogeny at interspecific levels, it does not seem to be a m eans of habitat partitioning by competing species of chewing lice that live on the same pocket gopher host (chapter 2). Certain species of chewing lice can coexist with other louse species, but this situation is the exception, rather than the rule in the gopher-louse system. For example, two genera of lice (Thomomydoecus an d Geomydoecus) occasionally co-occur on pocket gophers of the genus T h o m o m ys. In chapter 2, I showed that those coexisting louse species were spatially segregated over the pelage of the gopher, but the mechanism by which they partition their habitat does not appear to be differential ability to grasp hairs of different diameter. 105 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The last three chapters of this dissertation involve the search for new symbionts in the gopher-louse association, namely bacterial taxa associated with the chewing lice. In chapter 3, I used culture-independent methods pioneered by Carl Woese and Norm Pace to identify 35 bacterial lineages that were found in association with pocket gophers, chewing lice, and their subterranean burrow system. In chapter 4, I demonstrated that there was bacterial DNA associated with the eggs of chewing lice. Vertical transmission of endosymbionts through eggs is a common means of dispersal for endosymbiotic bacteria of insects. In the final chapter, I determined the phylogenetic relationships of two lineages of bacteria associated with this system. I sequenced the complete 16S rRINA gene for all clones related to the genera Staphylococcus an d Acinetobacter. I used parsimony and maximum likelihood analyses to generate phylogenetic hypotheses for the bacteria that were compared to those of their hosts (chewing lice). Unfortunately, little can be said about w hether bacteria show evidence of cophylogeny with their hosts until taxon sampling is more complete. It also will be necessary to determine whether thtese bacteria are true endosymbionts or merely transients in this system. Future research in this area will focus on determining which of the 35 bacterial lineages are truly endosymbiotic, further sampling of Staphylococcus a n d Acinetobacter, and identifying additional lineages of bacteria that almost certainly exist, but have yet to be discovered. 106 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LITERATURE CITED Aksoy, S., A. A. Pourhosseini, and A. Chow. 1995. M ycetome endosym bionts of tsetse flies constitute a distinct lineage related to Enterobacteriaceae. Insect Molecular Biology 4:15-22. Aksoy, S. 1995. Molecular analysis of the endosymbionts of tsetse flies: 16S rDNA locus and over-expression of a chaperonin.Insect Molecular Biology 4:23-29. Aksoy, S., X. Chen, and V. Hypsa. 1997. Phylogeny and potential transmission routes of midgut-associated endosymbionts of tsetse (Diptera: Glossinidae). Insect Molecular Biology 6:183-190. Baker, R. J., and S. L. Williams. 1972. A live trap for pocket gophers. Journal o f Wildlife Management 36:1320-1322. Bandi, C., G. Damiani, L. Magrassi, A. Grigolo, R. Fani, and L. Sacchi. 1994. Flavobacteria as intracellular symbionts in cockroaches. Procedings of the Royal Society of London B 257:43-48. Baumann, P., L. Baumann, C. Lai, and D. Rouhbakhsh, N. A. Moran, and M. A. Clark. 1995. Genetics, physiology, and evolutionary relationships of the genus Buchnera: Intracellular symbionts of aphids. Annual Review of Microbiology 49:55-94. Baumann, P. and N. A. Moran. 1997. Non-cultivable microorganisms from symbiotic associations of insects and other hosts. Antonie von Leeuwenhoek 72:39-48. Baumann, P., N. A. Moran, and L. Baumann. 1997. The evolution and genetics of aphid endosymbionts. Bioscience 47:12-20. Berchtold, M., W. Ludwig, and H. Kunig. 1994. 16S rDNA sequence and phylogenetic position of an uncultivated spirochete from the hindgut of the termite Mastotermes darwiniensis Froggatt. FEMS Microbiology Letters 123:269-274. Brooks, D. R. 1980. Allopatric speciation and non-interactive parasite community structure. Systematic Zoology 29:192-203. Brune, A., and M. Friedrich. 2000. Microecology of the termite gut: structure and function on a microscale. Current Opinion in Microbiology 3:263-269. Brynnel, E. U., C. G. Kurland, N. A. Moran, and S. G. E. Anderson. 1998. Evolutionary rates for tu f genes in endosymbionts of aphids. M olecular Biology and Evolution 15:574-582. 107 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Cary, S. C., W. Warren, E. Anderson, and S. J. Giovannoni. 1993. Identification and localization of bacterial endosymbionts in hydrothermal vent taxa with symbiont-specific polymerase chain reaction amplification and in situ hybridization techniques. Molecular Marine Biology and Biotechnology 2:51-62. Cunningham, C. W., H. Zhu, and D. M. Hillis. 1998. Best-fit maximum likelihood models for phylogenetic inference: empirical tests with known phylogenies. Evolution, 52:978-987. Dadd, R. H. 1985. N utrition: organisms, Pp.315-319, in, Comprehensive insect physiology, biochemistry, and pharmacology, vol. 4. (G. A. Kerkut and L. I. Gilbert, eds.). Pergamon Press, Inc., Oxford, N. Y. Demastes, J. W. and M. S. Hafner 1993. Cospeciation of pocket gophers and their chewing lice. Journal of Mammalogy 74:521-530. Demastes, J. W., M. S. Hafner, D. J. Hafner, and T. A. Spradling. 1998. Pocket gophers and chewing lice: a test of the maternal transmission hypothesis. Molecular Ecology 1065-1069. 7: Douglas, A. E. 1989. Mycetocyte symbiosis in insects. Biological Review Cambridge Philosophical Society 64:409-434. Douglas, A. E., and W. A. Prosser. 1992. Synthesis of the essential amino acid tryptophan in the pea aphid Acyrthosiphon( pisum) symbiosis. Journal o f Insect Physiology 38:565-568. Durden, L. A. 1987. Interspecific associations between epifaunistic arthropods on the eastern chipmunk, Tamias striatus, and an overview of arthropod coexistence on mammals in general. Journal of the Tennessee Academy of Science 62:44-51. Eberle, M. W., and D. L. McLean 1982. Initiation and orientation of the symbiote migration in the human body louse Pediculus humanus L. Journal of Insect Physiology 28:417-422. Eberle, M. W., and D. L. McLean 1983. Observation of symbiote migration in human body lice with scanning and transmission electron microscopy. Canadian Journal of Microbiology 29:755-762. Emerson, K. C., and R. D. Price. 1981. A host-parasite list of the Mallophaga on m am m als. Miscellaneous Publications of the Entomological Society of America 12:1-72. Farrelly, V., F. A. Rainey, and E. Stackebrandt. 1995. Effect of genomic size and rrn gene copy number on PCR amplification of 16S rRNA genes from a mixture of bacterial species. Applied and Environmental Microbiology 61:2798-2801. 108 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Felsenstein, J. 1981. Evolutionary trees from DNA sequences: a maximum likelihood approach. Journal of Molecular Evolution 17:368-376. Felsenstein, J. 1985. Phylogenies and the comparative method. Am erican Naturalist 125:1-15. Garland, T. Jr., P. H. Harvey, and A. R. Ives. 1992. Procedures for the analysis of comparative data using phylogenetically independent contrasts. Systematic Biology 41:18-32. Gause, G. F. 1934. The struggle for existence. The Williams and Wilkins Company, Baltimore, Maryland, 163 pp. Hafner, M. S. 1982. A biochemical investigation of geomyoid systematics (Mammalia: Rodentia). Z. Zool. Syst. Evolutions Forsch. 20:118-130. Hafner, M. S., and R. D. M. Page. 1995. M olecular phylogenies and host-parasite cospeciation. Philosophical Transactions of the Royal Society of London 349:77-83. Hafner, M. S., J. W. Demastes, T. A. Spradling, and D. L. Reed. In press. Cophylogeny between pocket gophers and chewing lice. In: (R.D.M. Page, ed.) Cophylogeny, Univ. Chicago Press. Hafner, M. S., and S. A. Nadler. 1988. Phylogenetic trees support the coevolution of parasites and their hosts. Nature 332:258-259. Hafner, M. S., and S. A. Nadler. 1990. Cospeciation in host-parasite assemblages: comparative analysis of rates of evolution and timing of cospeciation events. Systematic Zoology 39:192-204. Hafner, M. S., P. D. Sudman, F. X. Villablanca, T. A. Spradling, J. W. Demastes, and S. A. Nadler. 1994. Disparate rates of molecular evolution in cospeciating hosts and parasites. Science 265:1087-1090. Harvey, P. H., and A. E. Keymer 1991. Comparing life histories using phylogenies. Philosophical Transactions of the Royal Society of London Series 332:31- B 39. Harvey, P. H., and M. D. Pagel. 1991. The comparative method in evolutionary biology. Oxford University Press, Oxford, United Kingdom. 239 pp. Hellenthal, R. A., and R. D. Price 1984. Distributional associations among Geomydoecus and Thomomydoecus lice (Mallophaga: Trichodectidae) and pocket gopher hosts of the Thomomys bottae group (Rodentia: Geomyidae). Journal of Medical Entomology 21:432-446. 109 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Hellenthal, R. A., and R. D. Price. 1991. Biosystematics of the chewing lice of pocket gophers. Annual Review of Entomology 36:185-203. Hellenthal, R. A., and R. D. Price. 1994. Two new subgenera of chewing lice (Phthiraptera: Trichodectidae) from pocket gophers (Rodentia: Geomyidae), with a key to all included taxa. Journal of Medical Entomology 31:450-466. Honeycutt, R. L., and S. L. Williams. 1982. Genic differentiation in pocket gophers of the genus Pappogeomys, with comments on intergeneric relationships in the subfamily Geomyinae. Journal of Mammalogy 63:208-217. Hugenholtz, P., B. M. Goebel, and N. R. Pace. 1998. Impact of culture-independent studies on the emerging phylogenetic view of bacterial diversity. Journal o f Bacteriology 180:4765-4774. Hugenholtz, P., and N. R. Pace. 1996. Identifying microbial diversity in the natural environment: a molecular phylogenetic approach. Trends in Biotechnology 14:190-197. Hugenholtz, P., C. Pitulle, K. L. Hershberger, and N. R. Pace. 1998. Novel Division level bacterial diversity in a Yellowstone hot spring. Journal of Bacteriology 180:366-376. Ishikawa, H. 1989. Biochemical and molecular aspects of endosymbiosis in insects. International Review of Cytology 116:1-45. Johnson, K. P., and D. H. Clayton. In press. Cophylogeny at the community level: Comparing phylogenies of wing and body lice to that of columbiform hosts. In: (R.D.M. Page, ed.) Cophylogeny, Univ. Chicago Press. Kellogg, V. L. 1913. Distribution and species-forming of ectoparasites. The American Naturalist 47:129-158. Komatsoulis, G. A., and M. S. Waterman 1997. A new computational method for detection of Chimeric 16S rRNA artifacts generated by PCR Amplification from mixed bacterial populations. Applied and Environmental Microbiology 63:2338-2346. LaBarbera, M. 1989. Analyzing body size as a factor in ecology and evolution. Annual Review of Ecology and Systematics 20:97-117. Lambert, J. D., and N. A. Moran. 1998. Deleterious mutations destabilize ribosomal RNA in endosymbiotic bacteria. Proceedings of the National Academy of Sciences 95:4458-4462. Lambert, L. H., T. Cox, K, Mitchell, R. A. Rossello-Mora, C. Del Cueto, D. E. Dodge, P. Orkland, and R. J. Cano. 1998. Staphylococcus succinus sp. nov., isolated from Dominican amber. International Journal of Systematic Bacteriology 48:511-518. 110 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Marshall, A. G. 1981. The Ecology of Ectoparasitic Insects. Academic Press, London. 459 pp. Martin, R. D., and A. D. Barbour. 1989. Aspects of line-fitting bivariate allometric analyses. Folia Primatologica 53:65-81. Mathiak, H. A. 1938. A key to the hairs of the mammals of southern Michigan. The Journal of Wildlife Management 2:251-268. Mayer, W. V. 1952. The hair of California mammals with keys to the dorsal guard hairs of California mammals. The American Midland Naturalist 48:480-512. Moran, N. A. 1996. Accelerated evolution and Muller's ratchet in endosymbiotic bacteria. Proceedings of the National Academy of Sciences 93: 2873-2878. Moran, N. A., M. A. Munson, P. Baumann, and H. Ishikawa. 1993. A molecular clock in endosymbiotic bacteria is calibrated using the insect hosts. Proceedings of the Royal Society of London Series 253:167-171. B Moran, N. A., C. D. von Dohlen, and P. Baumann 1995. Faster evolutionary rates in endosymbiotic bacteria than in cospeciating insect hosts. Journal o f Molecular Evolution 41:727-731. Moran, N. A., and P. Baumann. 1994. Phylogenetics of cytoplasmically inherited microorganisms of arthropods. Trends in Ecology and Evolution 9:15-20. Moran, N. A., and A. Telang. 1998. Bacteriocyte-associated symbiosis of insects. Bioscience 48:295-304. Morand, S., M. S. Hafner, R. D. M. Page, and D. L. Reed. 2000. Comparative body size relationships in pocket gophers and their chewing lice. Biological Journal of the Linnean Society 70:239-249. Munson, M. A., P. Baumann, M. A. Clark, L. Baumann, N. A. Moran, D. J. Voegtlin, and B. C. Campbell. 1991. Evidence for the establishment of aphid-Eeubacterium endosymbiosis in an ancestor of four aphid families. Journal of Bacteriology 173:6321-6324. Munson, M. A., L. Baumann, and P. Baumann. 1993. Buchnera aphidicola (a prokaryotic endosymbiont of aphids) contains a putative 16S rRNA operon unlinked to the 23S rRNA-encoding gene: sequence determination, and promoter and terminator analysis. Gene 137:171-178. Murray, M. D. 1957a. The distribution of the eggs of M am m alian Lice on their hosts I. Description of the oviposition behaviour. Australian Journal of Zoology 5:13-18. I l l Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Murray, M. D. 1957b. The distribution of the eggs of Mammalian Lice on their hosts. H. Analysis of the oviposition behaviour of Damalma ovis (L.). Australian Journal of Zoology 5:19-29. Nardon, P., and A. M. Grenier. 1989. CRC Handbook on Endocytobiosis in Coleoptera: Biological, biochemical, and genetic aspects. Pp 175-216 in: Insect endocytobiosis: Morphology, physiology, genetics, and evolution. (W. Schwemmler and G. Gassner, eds.) 272pp. Nowak, R. M. 1999. Walker's mammals of the world. The Johns Hopkins University Press, Baltimore, Maryland 1629 pp. Ochman, H., S. Elwyn, and N. A. Moran. 1999. Calibrating bacterial evolution. Proceedings of the National Academy of Sciences 96:12638-12643. Ochman, H., and A. C. Wilson. 1987. Evolution in bacteria: evidence for a universal substitution rate in cellular genomes. Journal of M olecular Evolution 26:74-86. Ohkuma, M., and T. Kudo. 1996. Phylogenetic Diversity of the Intestinal Bacterial Community in the Termite Reticulitermes speratus. Applied and environmental microbiology 62:461-468. Pace, N. R. 1997. A molecular view of microbial diversity and the biosphere. Science 276:734-740. Page, R. D. M. 1990a. Component analysis: A valiant failure? Cladistics 6:119-136. Page, R. D. M. 1990b. Temporal congruence and cladistic analysis of biogeography and cospeciation. Systematic Zoology 39:205-226. Page, R. D. M. 1993a . COMPONENT 2.0. Tree comparison software for use with Microsoft® Windows™. The Natural History Museum, London, UK. Page, R. D. M. 1993b. Genes, organisms, and areas: the problem of multiple lineages. Systematic Biology 42:77-84. Page, R. D. M. 1994. Maps between trees and cladistic analysis of historical associations among genes, organisms, and areas. Systematic Biology 43:58- 77. Page, R. D. M. in press. Cospeciation. University of Chicago Press. Page, R. D. M., and M. S. Hafner. 1996. Molecular phylogenies and host-parasite cospeciation: gophers and lice as a model system. Pp. 255-270, in: New Uses for New Phylogenies (P. H. Harvey, et al., eds.). Oxford Univ. Press, 349 pp. 112 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Page, R. D. M., R. D. P. Price, and R. A. H ellenthal. 1995. Phylogeny of Geomydoecus and Thomomydoecus pocket gopher lice (Phthiraptera: Trichodectidae) inferred from cladistic analysis of adult and first-instar morphology. Systematic Entomology 20:129-143. Patton, J. L., and P. V. Brylski. 1987. Pocket gophers in alfalfa fields: causes and consequences of habitat-related body size variation. American Naturalist 130:493-506. Patton, J. L., M. F. Smith, R. D. Price, and R. A. Hellenthal. 1984. Genetics of hybridization between the pocket gophers Thomomys bottae and Thomomys tozonsendii in northeastern California. Great Basin Naturalist 44:431-440. Posada, D., and K. A. Crandall (1998). Modeltest: testing the model of DNA substitution. Bioinformatics 14:817-818. Price, R. D. and K. C. Emerson 1972. A new subgenus and three new species of Geomydoecus (Mallophaga: Trichodectidae) from Thom om ys (Rodentia: Geomyidae). Journal of Medical Entomology 9:463-467. Price, R. D., R. A. Hellenthal, and M. S. Hafner. 1985. The Geomydoecus (Mallophaga: Trichodectidae) from the Central American pocket gophers of the subgenus Macrogeomys (Rodentia: Geomyidae). Proceedings of the Entomological Society of Washington 87:432-443. Purcell, A. H., T. Steiner, F. Megraud, and J. Bove. 1986. In vitro isolation of a transovarially transmitted bacterium from the leafhopper Euscelidius variegatus (Hemiptera: Cicadellidae). Journal of Invertebrate Pathology 48:66-73. Purvis, A., and A. Rambaut. 1995. Comparative analysis by independent contrasts (CAIC): an Apple Macintosh application for analyzing comparative data. Computer Applications for Biosciences 11:247-251. Rainey, F. A., E. Lang, and E. Stackebrandt. 1994. The phylogenetic structure of the genus Acinetobacter. FEMS Microbiology Letters 124:349-354. Rainey, F. A., N. L. Ward-Rainey, P. H. Janssen, H. Hippe, and E. Stackebrandt. 1996. Clostridium paradoxum DSM 7308T contains multiple 16S rRNA genes with heterogeneous intervening sequences. Microbiology 142:2087- 2095. Reed, D. L. 1994. Possible mechanism for cospeciation in pocket gophers and lice, and evidence of resource partitioning among coexisting genera of lice. M. S. Thesis. Louisiana State University, Baton Rouge, Louisiana, 74 pp. 113 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reed, D. L., and M. S. Hafner. 1997. H ost specificity of chewing lice on pocket gophers: a potential mechanism for cospeciation. Journal of Mammalogy 78:655-660. Ries, E. 1931. Die symbiose der Lause und Federlinge. Zeitschrift fiir Morph. Okol. Tiere 20:233-367. Robison-Cox, J. F., M. M. Bateson, and D. M. Ward. 1995. Evaluation of nearest- neighbor methods for detection of chimeric small-subunit rRNA sequences. Applied and Environmental Microbiology 61:1240-145. Rust, R. W. 1974. The population dynam ics and host utilization of Geomydoecus oregonus, a parasite of Thomomys bottae. Oecologia (Berlin) 15:287-304. Sacchi, L., A. Grigolo, M. Mazzini, E. Bigliardi, B. Baccetti, and U. Laudani. 1988. Symbionts in the oocytes of Blattella germanica L. (Dictyoptera: Blattellidae): Their mode of transmission. International Journal of Insect Morphology and Embryology 17:437-446. Saffo, M. B. 1992. Invertebrates in endosymbiotic associations. Am erican Zoologist 32:57-565. SAS Institute Inc. 1994. SAS/STAT user's guide, release 6.03 edition. SAS Institute, Cary, North Carolina, 1,028 p. Saxena, A. K., and G. P. Agarwal. 1985. Mycetocytes in Aegypoecus perspicuus (Phthiraptera). Current Science 54:763-764. Saxena, A. K., G. P. Agarwal, S. C handra, and O. P. Singh. 1985. Pathogenic involvement of Mallophaga. Zeitshcrift fiir Angewandte Entomologie 99:294-301. Schilthuizen, M. and R. Stouthamer. 1997. Horizontal transmission of parthenogenesis-inducing microbes in Trichogramma wasps. Proceedings o f the Royal Society of London 264:361-365. B Silva, M. 1998. Allometric scaling of body length: elastic or geometric similarity in mammalian design. Journal of Mammalogy 79:20-32. Smith, D. C. and A. E. Douglas. 1987. The biology of symbiosis. Edward Arnold, London. 302 pp. Smith, M. F. 1998. Phylogenetic relationships and geographic structure in pocket gophers in the genus Thomomys. Molecular Phylogenetics and Evolution 9:1-14. Smith, M. F., and J. L. Patton 1988. Subspecies of pocket gophers: Causal bases for geographic differentiation in Thomomys bottae. Systematic Zoology 37:163- 178. 114 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Spradling, T. A. 1997. Relative rates of molecular evolution in rodents and their symbionts. Ph.D. dissertation, Louisiana State University, Baton Rouge. 174 pp. Staley, J. T. 1999. Bacterial Diversity: a time for place. American Society for Microbiology News 65:681-687.. Staley J. T., and J. J. Gosink. 1999. Poles apart: Biodiversity and biogeography of sea ice bacteria. Annual Review of Microbiology 53:189-215. Suzuki, M. T., and S. J. Giovannoni. 1996. Bias caused by template annealing in the amplification of mixtures of 16S rRNA genes by PCR. Applied and Environmental Microbiology 62:625-630. Swofford, D. L. 2000. PAUP*. Phylogenetic Analysis Using Parsim ony (‘and O ther Methods). Version 4. Sinauer Associates, Sunderland, Massachusetts. SYSTAT, Inc. 1992. Statistics, version 5.2 edition. SYSTAT, Inc. Evanston, Illinois. Tamura, K., and M. Nei. 1993. Estimation of the num ber of nucleotide substitutions in the control region of mitochondrial DNA in humans and chimpanzees. Molecular Biology and Evolution 10:512-526. Tanner, M. A., B. M. Goebel, M. A. Dojka, and N. R. Pace. 1998. Applied and Environmental Microbiology 64:3110-3113. Thompson, J. D., T. J. Gibson, F. Plewniak, F. Jeanm ougin, and D. G. Higgins. 1997. The ClustalX windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Research 24:4876-4882. Timm, R. M. 1983. Farenholz's Rule and resource tracking: A study of host- parasite coevolution. Pp. 225-266, in: Coevolution (M. Nitecki, ed.), Univ. Chicago Press. United States Centers for Disease Control. 1984. Controlling head lice. 2nd ed. United States Department of Health and Human Service, Centers for Disease Control, (CDC) 84-8397:1-15. Unterman, B. M., P. Baumann, and D. L. McLean. 1989. Pea aphid symbiont relationships established by analysis of 16S rRNAs. Journal of Bacteriology 171:2970-2974. Waage, J. K. 1979. The evolution of insect/vertebrate associations. Biological Journal of the Linnean Society 12:187-224. 115 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Wang, G. C.-Y., and Y. Wang. 1997. Frequency of formation of chimeric molecules as a consequence of PCR coamplification of 16S rRNA genes from mixed bacterial genomes. Applied and Environmental Microbiology 63:4645-4650. Ward, D. M. 1998. A natural species concept for prokaryotes. Current Opinions in Microbiology 1:271-277. Ward, D. M., M. J. Ferris, S. C. Nold, and M. M. Bateson. 1998. Natural view of microbial biodiversity within hot spring cyanobacterial mat communities. Microbiology and Molecular Biology Reviews 62:1-18. Werren, J. H., D. W indsor, and L. R. Guo. 1995. D istribution of Wolbachia am ong neotropical arthropods. Proceedings of the Royal Society of London. Series B 262:197-204. Woese, C. R. 1994. There must be a prokaryote somewhere: microbiology's search for itself. Microbiological Reviews 58:1-9. 116 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. APPENDIX: SPECIMENS EXAMINED Specimens examined: Specimens are deposited in the Museum of Natural Science, Louisiana State University (LSUMZ) and the Museum of Vertebrate Zoology, University of California, Berkeley (MVZ). Interordinal comparisons O rder Family Species Specimen number Sex Mass (g) Insectivora Soricidae Cryptotis parva LSUMZ 23789 U 4.0 T alpidae Scalopus aquations LSUM Z 6981 M 113.4 C hiroptera Pteropodidae Pteropus LSUMZ 17726 M 900b Vespertilionidae Pipistrellus hesperus LSUMZ 22041 M 3.5 Primates Cercopithecidae Colobus guereza LSUMZ 26271' F 8,000b Loris id ae Nycticebus LSUMZ 28897* M 650b Lagomorpha Leporidae Lepus alleni LSUMZ 13465 M 3,500b Ochotonidae Ochotona princeps LSUMZ 35909 F 129.9 Rodentia Heteromyidae Dipodomys ordii LSUMZ 25321 M 72 M uridae Baiomys taylori LSUMZ 4629 F 8.4 Sciuridae Sciurus niger LSUMZ 28476 F 1,000 Artiodactyla Bovidae Taurolragus oryx LSUM Z 36155* U 600,000b Bovidae Madoqua LSUM Z 36156* M 5,000b Hyracoidea Procaviidae Procavia capensis LSUM Z 34649* M 4,000b Perissodactyla T apiridae Tapirus bairdii LSUMZ 6977 M 250,000b Equidae Equus LSUMZ 36157* U 400,000b C arnivora M ustelidae Mustela frenata LSUMZ 28014 M 207 U rsidae Ursus arctos LSUMZ 36158* U 600,000b (appendix cont.) 117 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Intrafamilial comparisons Rodentia Geomyidae Geomys breviceps LSUMZ 31451 M 180 Geomyidae G. pinetus LSUMZ 23655 M 360 Geomyidae G. tropicalis LSUMZ 34345 M 235 Geomyidae Orthogeomys cavator LSUMZ 29253 M 875 Geomyidae O. cherriei LSUMZ 29539 M 455 Geomyidae O. heterodus LSUMZ 29265 M 620 Geomyidae O. underwoodi LSUMZ 28368 M 260 Geomyidae Thomomys bottae LSUMZ 35992 M 210 Geomyidae T. bulbivorus LSUMZ 31313 M 410 Geomyidae T. mazama LSUMZ 31398 F 87 Geomyidae T. monticola LSUMZ 31411 M 94 Geomyidae T. talpoides LSUMZ 34387 F 91 Geomyidae T. townsendii LSUMZ 31264 M 300 Geomyidae T. umbrinus LSUMZ 34362 M 115 Intrageneric comparisons Rodentia Geomyidae Thomomys bottae LSUMZ 35992 M 210 Geomyidae T. bulbivorus LSUMZ 31313 M 410 Geomyidae T. mazama LSUMZ 31398 F 87 Geomyidae T. monticola LSUMZ 31411 M 94 Geomyidae T. talpoides LSUMZ 34387 F 91 Geomyidae T. townsendii LSUMZ 31264 M 300 Geomyidae T. umbrinus LSUMZ 34362 M 115 Intraspecific comparisons Rodentia Geomyidae Thomomys bottae LSUMZ 35992 M 210 Geomyidae T. bottae (juvenile) LSUMZ 20936 M 37.5 Geomyidae T. b. perpes MVZ 166402 M 236 Geomyidae T. b. perpes MVZ 140497 M 96 Geomyidae T. b. perpallidus MVZ 170521 M 189 Geomyidae T. b. perpallidus M VZ 170565 M 200 Geomyidae T. b. perpallidus M VZ 80595 M 74.2 Geomyidae T. b. perpallidus MVZ 80606 M 89.0 Body mass estim ated from Nowak (1999). ’ Specimens not collected from the wild. 118 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. VITA David Lee Reed was bom in Charlotte, North Carolina, on 5 December, 1968. He attended school in Charlotte until 1987 when he graduated from East Mecklenburg High School. He attend the University of North Carolina, Wilmington where he received his bachelor of science degree in biological sciences in 1991. He then attended Louisiana State University and received his master of science degree in zoology in December, 1994. David then worked for two years as a biologist at The Florida Aquarium in Tampa, Florida. He is currently finishing his doctoral research in zoology at Louisiana State University and anticipates graduation in December, 2000. He has accepted a postdoctoral research position in the Department of Biology, at the University of U tah. 119 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. DOCTORAL EXAMINATION AND DISSERTATION REPORT Candidates David L. Reed Major Fields Zoology Title of Dissertations Host and Parasite Interactions Among Three Symbionts: Pocket Gophers, Chewing Lice, and Endosymbiotic Bacteria Approveds Major Professor Dean of tneMiraduate School EXAMINING COMMITTEE: 7$ _ - CQ - ______ Date of nation: September 29, 2000 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.