Upload
others
View
0
Download
0
Embed Size (px)
Citation preview
DNA-based Species Delimitation of the Agriculturally Important Genus, Ravinia (Diptera: Sarcophagidae)
A dissertation submitted to the
Graduate School
Of the University of Cincinnati
In partial fulfillment of the requirements for the degree of
Doctor of Philosophy
In the Department of Biological Sciences
of the McMicken College of Arts and Sciences
By
Evan Scott Wong
B.S., Biological Sciences with Honors, University of Cincinnati, September 2010
Committee chair: Ronald W. DeBry
ii
Abstract Species, and their constituent populations, are the units upon which evolution acts.
The ability to accurately identify and delimit species boundaries is crucial for research, but
is also important for public policy. Traditionally, species were identified and delimited
based on morphological characteristics, yet sometimes species are cryptic, meaning no
diagnostic morphological traits can be found. Consequently, DNA sequence data are
increasingly being used in the areas of species delimitation and identification. As a model to
explore species delimitation and identification I use members of the flesh fly family
Sarcophagidae, specifically the genus Ravinia. Using Bayesian inference and maximum
likelihood methods, I infer a molecular phylogeny, using mitochondrial DNA. Several
paraphyletic relationships were inferred among species of Ravinia that are discordant with
current morphological taxonomy. I investigate whether this conflict between genetic
lineages and morphological identification within the species Ravinia anxia and Ravinia
querula is due to introgressive gene flow, retained ancestral polymorphism, or indicative of
the presence of unrecognized species. Using a combination of concatenated phylogenetic
analyses, population genetic analyses, migration analyses, and coalescent methodologies I
delimit the species Ravinia anxia and Ravinia querula. I find three morphologically cryptic
lineages that comprise the current morphologically-identified species R. anxia and R.
querula. With the findings of this species delimitation research, I then re-evaluate the
phylogenetic relationships of Ravinia using increased sampling of both species and loci.
iii
Copyright © 2015
by
Evan S. Wong
All Rights Reserved
iv
v
Acknowledgments
The majority of specimen collection work and a minor portion of molecular
sequencing were supported by Grant No. 2005-DA-BX-K102 awarded by the National
Institute of Justice, Office of Justice Programs, U.S. Department of Justice to Ronald W.
DeBry (P.I.). The opinions, findings, and conclusion or recommendations expressed in this
publication are the author(s) and do not necessarily reflect the views of the Department of
Justice.
Additional molecular work and a portion of specimen collection work were funded
by multiple Wiemenn/Wendel/Benedict Summer Fellowships to Evan S. Wong. The
majority of molecular work was supported by the GSGA Summer Research Fellowship of
the Graduate Student Governance Association (University of Cincinnati). Specimen
collection in the state of Ohio was supported by The Ohio Biological Survey’s Small
Research Grants, awarded to Dr. Gregory A. Dahlem (P.I.) and Evan S. Wong (co-P.I.).
General laboratory costs and summer support was supported by Scholarship America:
Genzyme’s Chart Your Own Course national scholarship for people living with lysosomal
storage disorders; and University of Cincinnati’s Research Council SUMR Mentor program.
Conference and travel support was awarded by the Graduate Student Governance
Association for the public presentation of portions of this research.
I would like to thank my advisor, Dr. Ronald W. DeBry for his help and advice on this
project. I first came to the DeBry lab as a doe-eyed undergraduate student, with very little
laboratory experience and a small amount of critical thinking capability. Without Ron’s
support, this dissertation, my research, and many areas of my life would be very different
than they are today. I would also like to thank my wonderful committee (Drs. Gregory
vi
Dahlem, Bryan Carstens, Theresa Culley, and Ken Petren), who helped in the formation of
my ideas and their scientific soundness, and the many hours of their time that they gave
me. Specifically, I would like to thank Dr. Carstens, who graciously invited me to The Ohio
State University where I was able to learn under the guidance of Mr. Jordan Satler and Dr.
Carstens the specifics of the coalescent-based programs and migration analyses that were
used in this dissertation. I thank Dr. Theresa Culley for her constant availability, support,
and insight. Without fail, I could go to talk to Dr. Culley without an appointment and she
would freely give of her time. I thank Dr. Ken Petren for his sound mind and available ear
(Dr. Petren would listen to my rants about things such as chairs, desks, offices, and door
locks). I also thank Dr. Petren for his knowledge in evolutionary ecology which greatly
improved this research. Finally, I thank Dr. Gregory Dahlem who I am deeply indebted to
and without whom I would not have been able to start or complete this research. Dr.
Dahlem has been a constant mentor and is an amazing scientist from whom I have learned
so much. The many, many hours of collecting and the time spent on nature trails are
experiences that will stay with me for the rest of my life. Dr. Dahlem invested many hours
into me; teaching proper insect collection techniques and identifications.
I would also like to thank Dr. Trevor Stamper. Dr. Stamper was the first person that I
met in the DeBry Lab and to whom started me on the path of science. Dr. Stamper was a
great mentor who always was available to me for support. These acknowledgements are far
too short a place to put all my thanks. I also thank Dr. Dave Hauber who helped train me in
microbiology and who I consider a good friend. I thank Mrs. Stephanie Gierek who came to
the DeBry lab as an undergraduate researcher and later became a graduate student in our
lab. I only wish that she could have joined the lab sooner so I could have spent more time in
vii
the lab researching with her. I also thank the many undergraduate students I was able to
mentor over these past several years: Gabrielle Lombardo (2011-2012), Annie Roesler
(2011), Andrea Fuitem (2012-2013), Brandon Baker (2012-2013), Jessica Harrington
(2013), Stephanie Gierek (2014-2015), Dane Larson (2014-2015), and Jacqui Yarman
(2014-2015).
I thank the Department of Biological Sciences at the University of Cincinnati.
Specifically, I thank the office staff and support: Sandra McGeorge, Charmaine Mitchell,
Mary Wischer, Heather Wischer, Paul McKenzie, and Roger Ruff. I would like to thank the
wonderful Ms. Julie Stacey and Mrs. Ellen Cordonnier. Julie made teaching a blast and I am
very glad that I was able to teach Microbiology with her.
Finally, I want to thank my family and friends for their patience and unwavering
support during this time. My friend, Elizabeth Brown, was a constant source of support and
encouragement over these past five years. She saw me at the best of times and the worst of
times and accepted me for who I am. Thank you for being my life-long friend. To my
brothers, Elliot and Ean, thank you for your encouragement during this period of my life. I
thank you for being understanding and accepting. I thank my parents, Perry and Sandy
Wong, for their support. It has taken a while to get here, and I thank you for being there
along the way. I thank my Aunt Vivian who impressed upon me the life philosophy of:
“Good, Better, Best. I will not rest until my good is better, and my better is best.”
I dedicate this dissertation to two groups of people. The first group to whom I
dedicate this dissertation to are people living with rare genetic diseases. The second group
to whom I dedicate this dissertation to is the LGBTQ community. Both of these groups have
had a significant impact on my life and I am a better person because of them.
viii
Table of Contents
Abstract .................................................................................................................................................... ii
Acknowledgments.................................................................................................................................. v
Table of Contents ................................................................................................................................ viii
List of Tables ...........................................................................................................................................xi
List of Figures ....................................................................................................................................... xii
Chapter One: Introduction & Dissertation Organization ......................................................... 1
Chapter Two: Discordance between morphological species identification and mtDNA
in the flesh fly genus, Ravinia (Diptera: Sarcophagidae) ................................................ 14
2.1 Abstract .......................................................................................................................................................15
2.2 Introduction ...............................................................................................................................................16
2.3 Materials & Methods ..............................................................................................................................19
2.4 Results .........................................................................................................................................................27
2.5 Discussion...................................................................................................................................................29
2.6 Conclusions ................................................................................................................................................34
2.7 Tables ...........................................................................................................................................................37
2.8 Figures .........................................................................................................................................................41
2.9 References ..................................................................................................................................................43
ix
Chapter Three: Coalescent DNA-based species delimitation within the flesh fly genus
Ravinia Robineau-Desvoidy, 1863 (Diptera: Sarcophagidae) ...................................... 50
3.1 Abstract .......................................................................................................................................................51
3.2 Introduction ...............................................................................................................................................52
3.3 Materials & Methods ..............................................................................................................................56
3.4 Results .........................................................................................................................................................66
3.5 Discussion...................................................................................................................................................70
3.6 Conclusions ................................................................................................................................................75
3.7 Tables ...........................................................................................................................................................79
3.8 Figures .........................................................................................................................................................83
3.9 References ..................................................................................................................................................87
3.10 Supplemental Materials .....................................................................................................................98
Appendix S3.1 Primer Information and Amplification Procedure ..........................................98
Appendix S3.2 Best Model Selection and Partition Schemes ................................................. 101
Supplemental Tables .............................................................................................................................. 101
Supplemental Figures ............................................................................................................................ 107
Chapter Four: Phylogenetic relationships of the flesh fly genus Ravinia (Diptera:
Sarcophagidae) ........................................................................................................................... 125
4.1 Abstract .................................................................................................................................................... 126
4.2 Introduction ............................................................................................................................................ 127
4.3 Materials & Methods ........................................................................................................................... 130
x
4.4 Results ...................................................................................................................................................... 136
4.5 Discussion................................................................................................................................................ 139
4.6 Conclusions ............................................................................................................................................. 143
4.7 Tables ........................................................................................................................................................ 144
4.8 Figures ...................................................................................................................................................... 150
4.9 References ............................................................................................................................................... 159
4.10 Supplemental Materials .................................................................................................................. 165
Appendix S4.1 Best Model Selection and Partition Schemes ................................................. 165
Supplemental Tables .............................................................................................................................. 167
Supplemental Figures ............................................................................................................................ 171
Chapter Five: General Conclusions ...................................................................................................... 173
xi
List of Tables
Chapter Two
Table 2.1. Specimen locality data and references for COI and COII sequences
included in this study. ..........................................................................................................37
Table 2.2. Tamura-Nei + G corrected distances within and between indicated
groups. ......................................................................................................................................40
Chapter Three
Table 3.1. Gene information for multilocus analyses. ..........................................................................79
Table 3.2. *BEAST marginal likelihoods. ..................................................................................................80
Table 3.3. Migration results using Migrate-n. ........................................................................................81
Table 3.4. Mean estimates of population parameters from IMa2 analyses. ................................82
Table S3.1. Primer sequence information. ...............................................................................................98
Table S3.2. List of specimens, specimen identification (I.D.) codes, collection data, and
GenBank Accession numbers (Chapter 3). ................................................. 104
Table S3.3 BP&P results with guide tree parameters. ..................................................................... 106
Chapter Four
Table 4.1. List of specimens, specimen identification (I.D.) codes, sequenced genes
and GenBank accession numbers (Chapter 4) ........................................................ 144
Table 4.2. Gene information in multilocus analyses. ......................................................................... 149
Table S4.1. Specimen sex and locality data... ........................................................................................ 167
xii
List of Figures
Chapter Two
Figure 2.1. The inferred phylogenetic hypothesis of Ravinia species relationships
of North America. .................................................................................................................41
Chapter Three
Figure 3.1. Maximum likelihood tree, inferred using RAxML, of the concatenated
(2 mtDNA loci + 5 nuDNA loci) data set for all samples. .......................................83
Figure 3.2. Results from species delimitation analyses, represented on BEAST tree
from bGMYC of all sampling localities. .........................................................................85
Figure S3.1. Inferred COI gene-tree using maximum likelihood and Bayesian
inference. ............................................................................................................................... 107
Figure S3.2. Inferred COII gene-tree using maximum likelihood and Bayesian
inference ................................................................................................................................ 109
Figure S3.3. Inferred EF-1a gene-tree using maximum likelihood and Bayesian
inference ................................................................................................................................ 111
Figure S3.4. Inferred White gene-tree using maximum likelihood and Bayesian
inference. ............................................................................................................................... 113
Figure S3.5. Inferred Period gene-tree using maximum likelihood and Bayesian
inference ................................................................................................................................ 115
Figure S3.6. Inferred G6PD gene-tree using maximum likelihood and Bayesian
inference ............................................................................................................................... 117
Figure S3.7. Inferred PEPCK gene-tree using maximum likelihood and Bayesian
inference ................................................................................................................................ 119
xiii
Figure S3.8. bGMYC analysis heat map and topology. ...................................................................... 121
Figure S3.9. *BEAST tree selected as the most supported species delimitation
scenario ................................................................................................................................... 123
Chapter Four
Figure 4.2. Sampling map of Ravinia species used for phylogenetic analyses in
this study ............................................................................................................................... 150
Figure 4.2. Inferred phylogenetic tree from concatenated 12 loci using the program
RAxML ...................................................................................................................................... 152
Figure 4.3. Wing diagram showing location/absence of setae on Chaetoravinia and Ravinia
individuals .............................................................................................................................. 155
Figure 4.4. Species-tree estimation from five mitochondrial and seven nuclear loci
using the program *BEAST.. ............................................................................................ 157
Figure S4.1. Species-tree estimation from seven nuclear loci using the program
*BEAST.. ................................................................................................................................... 171
1
Chapter One:
Introduction & Dissertation Organization
2
INTRODUCTION
Species delimitation is not just an abstract problem in biology, but is a crucial first
step for many kinds of biological research and a vitally important part of the relationship
between science and public policy. For example, the ability to accurately define a species is
fundamental for conservation (e.g., Lowenstein et al. 2009) and control of pest and invasive
species (e.g., Spillings et al. 2009). Indeed, delimiting species boundaries and determining
phylogenetic relationships have been described as the primary goals of systematics (Chen
et al. 2014). Wiens (2007) describes species delimitation as the process through which
species boundaries are determined, new species are discovered, and currently recognized
species are found to be synonymous. Despite considerable debate, there appears to be
consensus that species are independently evolving metapopulation lineages (de Queiroz
2007). The variation among different species concepts can be attributed to the point in
time on the speciation continuum when a lineage becomes recognized as a species. This
provides a justification for the use of genetic data as the primary data for delimiting
species, as genetic data reflect the historical processes that give rise to these independent
metapopulation lineages (i.e., species). Consequently, DNA sequence data are increasingly
being used in species delimitation studies, yet few studies utilize DNA data as the primary
tool in species delimitations.
Traditionally, morphological characteristics are the first and primary tools used in
the delimitation of species. In such situations, highly trained scientists conduct
identifications and develop keys of hypothesized characteristics that could be used to
distinguish between species. The presence of diagnostic characters can provide evidence
indicating that there is an absence of (or perhaps minimal) gene flow between species and
3
other populations in the environment (Wiens and Servedio 2000). However, there are
certain limitations to this methodology that include the possibility that no distinguishing
factor can be provided from morphological data alone, for example in cryptic species
(Hebert et al. 2004a, 2004b; Lowenstein et al. 2009; Spillings et al. 2009). However, even
when morphological characters successfully delimit species, the use of genetic data can
speed the process (Sites and Marshall 2004). Furthermore, using an integrative species
delimitation approach (e.g., morphology and genetic data) pushes taxonomy beyond
conventional naming to understanding the evolutionary processes that bring species about
(Schlick-Steiner et al. 2010).
With the efficient and rapid availability of DNA sequence data there has been an
increase in new methods used in species delimitation. One of the most prominent
approaches is the use of DNA sequence data in phylogenetic analyses, followed by
determination of species boundaries from the taxonomic status and phylogenetic
relationships (Catleka and McAllister 2004). Typically these methods have utilized
arbitrary cutoffs as indicators of species status such as sequence divergence and migration
rates. For example, the “10x rule” proposed by Hebert et al. (2004b) requires between-
species divergence to be at least 10 times that as seen within species. Other methods, based
upon the genealogical species concept (Baum and Shaw 1995), require that gene-trees at
all loci exhibit reciprocal monophyly. This imposes an unrealistic requirement on
speciation because of the very long time required to achieve reciprocal monophyly at
neutral loci (Hickerson et al. 2006). In addition, conflicting gene-trees are a natural
phenomenon and a common result of stochastic changes in the coalescent, or the alleles of
4
a gene shared by all members in the population, of ancestral species (Hudson and Coyne
2002; Rannala and Yang 2003).
Some recent examples of species delimitation have used mitochondrial DNA
(mtDNA) to infer species limits (Serb et al. 2001; Wiens and Penkrot 2002; Ballard and
Whitlock 2004). Mitochondrial DNA has been shown to be beneficial in species delimitation
studies, particularly in species discovery studies using a single locus (e.g., barcodes). One
useful property of mtDNA is that the rate at which haplotypes of smaller effective
population sizes (Ne) coalesce is four times faster than nuclear markers (Moore 1995). As a
result, newly formed species generally will become reciprocally monophyletic for their
mtDNA sequences before most nuclear markers (Wiens and Penkrot 2002), although this is
not always the case (Hudson and Turelli 2003). The expected time to coalescence assumes
that the sampled haplotypes are adaptively neutral and will depend on the genetic data
used (i.e., autosomal loci, mitochondrial loci, allosomal loci; Moore 1995).
The exclusive use of mtDNA for the discovery, validation, and identification of a
species is an area of controversy and it has been convincingly argued that species should
not be delimited based on these data alone (Moritz et al. 1992; Moritz 1994; Sites and
Crandall 1997). The mitochondrial genome produces only a single gene-tree regardless of
the number of base pairs sequenced (Wiens and Penkrot 2002; Amaral et al. 2010). Thus,
independent replicated data cannot be obtained regardless of how many genes are
sequenced. In addition, mtDNA has the propensity to introgress across species boundaries
when nuclear DNA (nuDNA) does not (Ballard and Whitlock 2004). This problem becomes
significant at the early stages of speciation, where both morphological and molecular
markers will likely show low magnitudes of divergence across species (de Queiroz 2007).
5
Problems that make the delimitation of recently diverged species difficult include
incomplete lineage sorting (ILS) and current gene flow (i.e., migration) between
populations or species (Maddison 1997; Hudson and Coyne 2002; Degnan and Rosenberg
2006). Although gene flow and ILS may lead to a similar topology on the inferred trees,
these two processes have different implications in species delimitation (Harrington and
Near 2012). Gene flow produces polyphyly in the gene-tree through the intrusion of allelic
diversity across species boundaries. Introgressive hybridization, as a result of gene flow
from one species into the gene pool of another, may lead to perceived genetic similarity
among otherwise phenotypically divergent individuals (Keck and Near 2010).
ILS is a natural phenomenon and results from a stochastic sorting of ancestral
alleles. ILS causes gene genealogies to vary across the genome, resulting in a gene-tree that
is often different from the actual species phylogeny (Pamilo and Nei 1988; Takahata 1989;
Hobolth et al. 2011). At speciation, and a considerable time after, the random sorting of
ancestral alleles will cause some daughter species to possess alleles that are more closely
related to a sister species, than to conspecifics. Thus, daughter species are expected to
produce polyphyletic gene-trees for some time after the speciation event. Eventually, the
diversity in allelic lineages will be lost through genetic drift resulting in a single allelic
lineage that is representative of each daughter species. However, this progression from
polyphyly to monophyly in the gene-trees can take a long amount of time (i.e., 4Ne
generations; Moore 1995).
Species delimitation using genetic data (nuclear and mitochondrial loci)
necessitates that one consider both the process of species divergence and the contribution
of stochastic genetic processes that may lead to discordance among gene-trees and species
6
boundaries (Knowles and Carstens 2007). It cannot be determined if an inferred
phylogeny, based upon a single locus, is experiencing introgression or ILS unless it is
analyzed with other independent loci. These problems encountered when exclusively using
mtDNA highlight the need to use multiple nuclear loci in species delimitation studies
(Ballard and Whitlock 2004). When multiple nuclear and mitochondrial loci are analyzed
together more consistent and accurate relationships can be inferred when using DNA-
based methods (Wang et al. 1997).
Study system: Ravinia (Diptera: Sarcophagidae)
Ravinia provides a model genus for exploration into the aspects of species
delimitation using DNA-based techniques. This study system was chosen because of its: 1)
broad geographic and ecological range (Wong et al. 2015), 2) importance in other fields
(i.e., agriculture; Pickens 1981), 3) longstanding taxonomic disagreements (e.g., Hall 1928;
Aldrich 1930; Dodge 1956), and 4) ambiguous species limits of some members based on
morphology (Wong et al. 2015).
The genus Ravinia is a common group of primarily coprophagous flies; as larvae
they consume and digest feces of animals (Wong et al. 2015). Currently, there are 34
recognized species worldwide, of which 17 occur in North America, 16 are Neotropical, and
1 species is from the Old World (Pape 1996; Pape and Dahlem 2010). Many species of
Ravinia are closely associated with human activity and have a direct impact on the
environment through their role in vertebrate waste decomposition. In addition, Ravinia
species have been investigated for their potential use as biological control agents, having
been shown to be effective against introduced pests such as the face fly, Musca autumnalis
(De Geer), and the horn fly Haematobia irritans (Linnaeus) (Pickens 1981; Thomas and
7
Morgan 1972). While Ravinia may be among the most common of all sarcophagid flies
collected in the U.S. based on its presence in institutional collections, little is still known
about the biology and evolution of many species within this genus. As of yet, there has been
no phylogenetic study that establishes relationships for species within this genus.
DISSERTATION ORGANIZATION
Chapter Two presents a phylogeny of the genus Ravinia and compares the recovered
lineages with the morphologically identified species. I analyze data from two mitochondrial
genes in order to better understand the phylogenetic relationships among species in this
genus. Using both Bayesian inference and maximum likelihood methods, I estimate the
mitochondrial phylogeny and find several novel relationships. Among species of Ravinia, I
find several highly supported paraphyletic relationships. However, this conflict between
the morphological species definitions and the mtDNA phylogeny could be indicative to the
presence of cryptic species, due to introgression, a result of incomplete lineage sorting, or a
combination of the three. To further investigate the causes of this discordance a species
delimitation study using multilocus data would need to be conducted.
In Chapter Three, I conduct a species delimitation study focusing on two species
within the genus Ravinia: Ravinia anxia (sensu lato) and Ravinia querula (sensu lato). Wong
et al. (2015) have shown that these species demonstrate discordance between the mtDNA
phylogeny and the morphologically identifications. However, this discordance could be the
result of incomplete lineage sorting, introgression and migration, or the presence of
unrecognized species. In order to understand the cause of this discordance, I delimit these
two species using a multilocus data set and several different species delimitation methods,
8
including migration analyses, population genetic analyses, coalescent-based species-tree
analyses, and standard phylogenetic analyses.
Chapter Four brings these two previously mentioned studies together, and using a
multilocus data set (5 mtDNA + 7 nuDNA loci) I investigate the phylogenetic status of the
three cryptic species recovered from the species delimitation of R. anxia and R. querula (R.
anxia (sensu stricto), R. querula (sensu stricto), Ravinia n.sp.), and the status of R. floridensis
and R. lherminieri. Bayesian inference, maximum likelihood, and coalescent-based species-
tree methods using these markers strongly supported the distinctiveness of each of the
Ravinia species, including showing the three cryptic species as monophyletic lineages. This
study is the first to employ a coalescent-based species-tree approach within the
Sarcophagidae using a large data set.
The final chapter, Chapter Five: General Conclusions, provides a synopsis of the
findings of the three studies found in Chapters Two, Three, and Four. In Chapter Five, I also
discuss the potential for future work within Ravinia and suggest possible avenues for
future research in this flesh fly genus.
9
REFERENCES
Aldrich, J.M. (1930). Notes on the types of American two-winged flies of the genus
Sarcophaga and a few related forms, described by early authors. Proc. U.S. Nat. Mus.
78:1-39.
Amaral, A.R., Silva, M.C., Möller, L.M., Beheregaray, L.B., Coelho, M.M. (2010). Anonymous
nuclear markers for cetacean species. Conserv. Genet. 11:1143-1146.
Ballard, J.W.O., Whitlock, M.C. (2004). The incomplete natural history of mitochondria. Mol.
Ecol. 13:729-744.
Baum, D.A., Shaw, K.L. (1995). Genealogical perspectives on the species problem. In: Hoch,
P.C., Stephenson, A.G., editors. Molecular and experimental approaches to plant
biosystematics. St. Louis, MO: Missouri Botanical Garden. p. 289-303.
Catleka, B.C., McAllister, B.F. (2004). A genealogical view of chromosomal evolution and
species delimitation in the Drosophila virilis species subgroup. Mol. Phyl. Evol.
33:664-670.
Chen, X., Jiang, K., Guo, P., Huang, S., Rao, D., Ding, L., Takeuchi, H., Che, J., Zhang, Y., Myers,
E.A., Burbrink, F.T. (2014). Assessing species boundaries and the phylogenetic
position of the rare Szechwan ratsnake, Euprepiophis perlacea (Serpentes:
Colubridae), using coalescent-based methods. Mol. Phyl. Evol. 70:130-136.
Degnan, J.H., Rosenberg, N.A. (2006). Discordance of species trees with their most likely
gene trees. PLoS Genetics 2:e68.
de Queiroz, K. (2007). Species concepts and species delimitation. Syst. Biol. 56:879-886.
Dodge, H.R. (1956). New North American Sarcophagidae, with some new synonymy
(Diptera). Ann. Entomol. Soc. Am. 49:182-190.
10
Hall, D.G. (1928). Sarcophaga pallinervis and related species in the Americas. Ann. Entomol.
Soc. Am. 21:331-352.
Harrington, R.C., Near, T.J. (2012). Phylogenetic and coalescent strategies of species
delimitation in snubnose darters (Percidae: Etheostoma). Syst. Biol. 61:63-79.
Hebert, P.D.N., Penton, E.H., Burns, J.M., Janzen, D.H., Hallwachs, W. (2004a). Ten species in
one: DNA barcoding reveals cryptic species in the neotropical skipper butterfly
Astraptes fulgerator. Proc. Natl. Acad. Sci. USA 101:14812-14817.
Hebert, P.D.N., Stoeckle, M.Y., Zemlak, T.S., Francis, C.M. (2004b). Identification of birds
through DNA barcodes. PLoS Biol. 2:e312.
Hickerson, M.J., Meyer, C.P., Moritz, C. (2006). DNA barcoding will often fail to discover new
animal species over broad parameter species. Syst. Biol. 55:729-739.
Hobolth, A., Dutheil, J.Y., Hawks, J., Schierup, M.H., Mailund, T. (2011) Incomplete lineage
sorting patterns among human, chimpanzee, and orangutan suggest recent
orangutan speciation and widespread selection. Genome Res. 21:349-356.
Hudson, R.R., Coyne, J.A. (2002). Mathematical consequences of the genealogical species
concept. Evolution 56:1557-1565.
Hudson, R.R., Turelli, M. (2003). Stochasticity overrules the “Three-times rule”: genetic
drift, genetic draft, and coalescence times for nuclear loci versus mitochondrial DNA.
Evolution 57:182-190.
Keck, B.P., Near, T.J. (2010). Geographic and temporal aspects of mitochondrial
replacement in Nothonotus darters (Teleostei: Percidae: Etheostomatinae).
Evolution 63:1410-1428.
11
Knowles, L.L., Carstens, B.C. (2007). Delimiting species without monophyletic gene trees.
Syst. Biol. 56:887-895.
Lowenstein, J.H., Amato, G., Kolokontronis, S.-O. (2009). The real maccoyii: identifying tuna
sushi with DNA barcodes – constraining characteristic attributes and genetic
distances. PLoS ONE. 4:e7866.
Maddison, W.P. (1997). Gene trees in species trees. Syst. Biol. 58:289-311.
Moore, W.S. (1995). Inferring phylogenies from mtDNA variation: mitochondrial gene trees
versus nuclear-gene tree. Evolution. 49:718-726.
Moritz, C. (1994). Application of mtDNA in conservation: a critical review. Mol. Ecol. 3:401-
411.
Moritz, C., Schneider, C.J., Wake, D.B. (1992). Evolutionary relationships within the Ensatina
escholtzii complex confirm the ring species interpretation. Syst. Biol. 41:273-291.
Pamilo, P., Nei, M. (1988). Relationships between gene trees and species trees. Mol. Biol.
Evol. 5:568-583.
Pape, T. (1996). Genus Ravinia Robineau-Desvoidy. In: Catalogue of the Sarcophagidae of
the world (Insecta: Diptera). Gainesville, FL, USA, Associated Publishers. p 284-290.
Pape, T., Dahlem, G.A. (2010). Sarcophagidae (Flesh Flies). In: Brown, B.V. et al., editors.
Manual of Central American Diptera, Volume 2. Ottawa, Ontario, CAN. p. 1313-1335.
Pickens, L.G. (1981). The life history and predatory efficiency of Ravinia lherminieri
(Diptera: Sarcophagidae) on the face fly (Diptera: Muscidae). Canad. Entomol.
113:523-526.
Rannala, B., Yang, Z. (2003). Bayes estimation of species divergence times and ancestral
population sizes using DNA sequences from multiple loci. Genetics 164:1645-1656.
12
Schlick-Steiner, B.C., Steiner, F.M., Seifert, B., Stauffer, C., Christian, E., Crozier, R.H. (2010).
Integrative taxonomy: a multisource approach to exploring biodiversity. Annu. Rev.
Entomol. 55:421-438.
Serb, J.M., Phillips, C.A., Iverson, J.B. (2001). Molecular phylogeny and biogeography of
Kinosternon flavescens based on complete mitochondrial control region sequences.
Mol. Phylo. Evol. 18:149-162.
Sites, J.W., Jr., Marshall, J.C. (2004). Operational criteria for delimiting species. Annu. Rev.
Ecol. Syst. 35:199-227.
Spillings, B.L., Brooke, B.D., Koekemoer, L.L., Chiphwanya, J., Coetzee, M., Hunt, R.H. (2009).
A new species concealed by Anopheles funestus Giles, a major malaria vector in
Africa. Am. J. Trop. Med. Hyg. 81:510-515.
Takahata, N. (1989). Gene genealogy in three related populations: consistency probability
between gene and population trees. Genetics. 122:957-966.
Thomas, G.D., Morgan, C.E. (1972). Parasites of the horn fly in Missouri. J. Econ. Entomol.
65:169-74.
Wang, R.L., Wakely, J., Hey, J. (1997). Gene flow and natural selection in the origin of
Drosophila pseudoobscura and close relatives. Genetics. 147:1091-1106.
Wiens, J.J. (2007). Species delimitation: new approaches for discovering diversity. Syst. Biol.
56:875-878.
Wiens, J.J., Penkrot, T.A. (2002). Delimiting species based on DNA and morphological
variation and discordant species limits in spiny lizards (Sceloporus). Syst. Biol.
51:69-91.
13
Wiens, J.J., Servedio, M.R. (2000). Species delimitation in systematics: inferring diagnostic
differences between species. Proc. R. Soc. Lond. [Biol]. 267:631-636.
Wong, E.S., Dahlem, G.A., Stamper, T.I., DeBry, R.W. (2015). Discordance between
morphological species identification and mtDNA phylogeny in the flesh fly genus
Ravinia (Diptera: Sarcophagidae). Invert. Syst. 29:1-11.
14
Chapter Two:
Discordance between morphological species identification and mtDNA in the
flesh fly genus Ravinia (Diptera: Sarcophagidae).1
1 This article is the pre-publication version of the published article: Wong, E.S., Dahlem, G.A., Stamper, T.I., & DeBry, R.W. (2015) Discordance between
morphological species identification and mtDNA phylogeny in the flesh fly genus Ravinia (Diptera: Sarcophagidae). Invertebrate Systematics 29, 1-11.
The definitive version of this article can be accessed via Invertebrate Systematics. <http://dx.doi.org/10.1071/IS14018>
15
2.1 ABSTRACT
In order to better understand the phylogenetic relationships among species in the
genus Ravinia Robineau-Desvoidy, 1863, we analyzed data from two mitochondrial gene
fragments: cytochrome oxidase I (COI) and cytochrome oxidase II (COII). We used Bayesian
inference and maximum likelihood methods to infer phylogenetic relationships. Our results
indicate that the genera Ravinia and Chaetoravinia, previously synonymized into the genus
Ravinia (sensu lato) are each likely to be monophyletic (posterior probability 1; bootstrap
support 85%). We found highly supported paraphyletic relationships among species of
Ravinia, with relatively deep splits in the phylogeny. This conflict between the
morphological species definitions and the mtDNA phylogeny could be indicative of the
presence of cryptic species in Ravinia anxia (Walker, 1849), Ravinia floridensis (Aldrich,
1916), Ravinia lherminieri (Robineau-Desvoidy, 1830), and Ravinia querula (Walker, 1849).
Keywords: COI, COII, Chaetoravinia, mitochondrial DNA, molecular, morphology,
phylogenetics, North America
16
2.2 INTRODUCTION
The genus Ravinia Robineau-Desvoidy, 1863 represents an agriculturally important
lineage of flesh flies in the Dipteran family Sarcophagidae. Ravinia species are primarily
coprophagous flies that feed on a variety of mammalian species’ dung, including common
livestock animals such as cattle, horses, and pigs. This genus has a broad distribution with
34 species recognized worldwide: 1 Old World species and 33 species with a New World
distribution (Pape 1996; Pape and Dahlem 2010). Ravinia may be among the most common
of all flesh flies collected in North America, based on the abundant presence of these flies in
insect collections. Although larvae of Ravinia species are believed to be primarily
coprophagous, R. anxia (Walker, 1849) has been shown to be a facultative predator of the
face fly, Musca autumnalis De Geer, 1776, larvae in bovine feces (Pickens 1981).
Interestingly, Ravinia do not pose a nuisance to humans. They do not normally enter
houses and other human built structures and have little interest in landing on exposed skin.
These flies are not known to drink livestock eye secretions and are not attracted to
wounds. As such, an increase in Ravinia populations could have a beneficial effect for both
urban areas and rural farms in the reduction of accumulated mammalian waste.
The currently defined genus, Ravinia, has been separated into three genera at
various times and by various authors: Ravinia, Adinoravinia Townsend, 1917, and
Chaetoravinia Townsend, 1917 (Lopes 1969). These groups have since been fused into the
modern Ravinia (sensu lato) (Pape 1996). To date, no phylogenetic relationships of species
within this genus have been proposed. Ravinia has proven to be a taxonomically difficult
group, in that member species exhibit limited numbers of diagnostic morphological
characters. Robineau-Desvoidy (1863) originally established Ravinia based on the
17
hemispherical shape of the abdomen, with no hairs on the first segments of the abdomen
(female), and lack of tufts of hairs on the hind tibia (male). However, these types of
characters cannot be used to distinguish modern sarcophagid genera. Adults of the genus
Ravinia have the following identifying characteristics: postalar wall setulate; tegulae
orangish in ground color; frontal vitta of female at midpoint more than two times the width
of parafrontal plate; male lacking apical scutellar bristles and with a reddish-orange genital
capsule.
Morphological characteristics, in general, are the first and primary tools used in the
identification of species. In the modern context, trained taxonomists carry out
identifications and develop keys of diagnostic character traits that could be used to
distinguish between species. Some argue that the presence of cryptic species limits the
ultimate utility of morphological taxonomy (Hebert et al. 2004a, 2004b; Spillings et al.
2009). The identification of diagnostic characters is important because it indicates that
there is an absence of (or perhaps minimal) gene flow between species (Wiens and
Servedio 2000). The degree of intra-specific variation in morphological traits present in
Ravinia has made it difficult to make precise and accurate identifications of species
members. This limited number of distinctive species-level morphological characters makes
evolutionary reconstruction of species relationships difficult. Integrating DNA sequence
data into taxonomy provides new methods for species identification.
One of the most prominent approaches is the use of DNA sequence data in analyses
of taxonomic status and phylogenetic relationships (e.g., Catleka and McAllister 2004).
DNA-based approaches are likely to be particularly important in groups like Ravinia that
are characterized by limited numbers of discriminatory morphological characteristics at
18
the species-level. The ability to group individuals into separate genetic groups using DNA
sequence data may allow for the discovery of new morphological characters that would
enable easier and more reliable separation of these species by morphological means.
Mitochondrial DNA (mtDNA) sequence data are useful tools for species discovery
and identification studies (Hebert et al. 2004a, 2004b; Smith et al. 2007; Reid and Carstens
2012). Mitochondrial DNA is easily obtained due to the high abundance in cells relative to
nuclear DNA (Wells and Stevens 2009) and much of the sequence data available for the
class Insecta is for mitochondrial genes (Caterino et al. 2000). Currently, the vast majority
of described species are non-model organisms where in-depth molecular data is lacking. As
a result, molecular markers are often chosen based on their use in previous molecular
studies, not necessarily because they are the most adequate. One useful property of mtDNA
is that with smaller effective populations sizes (Ne) the haplotypes will coalesce four times
faster than nuclear markers (Moore 1995). Consequently, newly formed species will
become distinct in their mtDNA sequences before they become distinct for most nuclear
markers (Wiens and Penkrot 2002). This time to coalescence, however, assumes that the
sampled haplotypes are adaptively neutral and, thus, will depend on the markers chosen
(Moore 1995).
To understand if disagreement exists between the mitochondrial genealogy and the
morphological characteristics currently used in species identification, we inferred a
phylogenetic tree using two mitochondrial loci: cytochrome oxidase I (COI) and
cytochrome oxidase II (COII). We also discuss the presence of two distinct clades,
Chaetoravinia and Ravinia, within the genus Ravinia (sensu lato) and suggest future studies
to determine if subgeneric or generic status is warranted.
19
2.3 MATERIALS AND METHODS
Specimen collection and sampling
Flies were collected by net, Malaise trap, and bait traps using dog dung or aged
chicken thigh and liver as an attractant. Specimens were frozen in bulk and taken to the lab
for morphological identification by GAD and molecular analysis (based on Stamper et al.
2013). DNA was extracted from 1-3 legs for use in analyses, thereby preserving all
morphological structures important for species identification. Legs were removed and
transferred to a vial of 95% EtOH and stored at -80C until DNA extraction. The remainder
of the specimen was then pinned or stored in a separate vial of EtOH at -80C. All
specimens used in this study are retained as vouchers (currently held by GAD, these will be
placed in an established natural history collection upon completion of current studies).
Kutty et al. (2010) found the genus Blaesoxipha Loew, 1861 to be sister to Ravinia,
while other analyses, particularly those based on larval and adult morphological
characters, have more often supported the genus Oxysarcodexia Townsend, 1917 as being
the sister to Ravinia (Lopes 1982; Pape 1994; Giroux et al. 2010). A recent analysis based
on two mitochondrial genes, COI and COII, found Oxysarcodexia to be the closest relative to
Ravinia (Stamper et al. 2013). Therefore, we included in this study four outgroups, two
from each of these genera: Oxysarcodexia ventricosa (Wulp, 1895); O. cingarus (Aldrich,
1916); Blaesoxipha arizona Pape, 1994; B. cessator (Aldrich, 1916).
We sampled several individuals from multiple populations for 10 of the 17 species
of Ravinia in North America. We were not successful in obtaining material for the following
species, despite collection efforts in regions where they have been reported to occur:
Ravinia acerba (Walker, 1849), R. anandra (Dodge, 1956); R. coachellensis (Hall, 1931); R.
20
effrenata (Walker, 1861); R. pectinata (Aldrich, 1916); R. sueta (van der Wulp, 1895); and R.
tancituro Roback, 1952. Individuals representing two of the three previously synonymized
genera, Chaetoravinia and Ravinia were included in this study. The specimens included in
this study, their locality information, and GenBank accession numbers can be found in
Table 2.1.
Nomenclatural Notes
The nomenclature involved with R. anxia has been rather confused in the past and
much of the confusion deals with the separation of this species from R. querula and the
mistaken identity of R. lherminieri. The following nomenclatural history of this species
should illustrate some of the problems involving the morphological separation of species in
this group of Ravinia. The first name that was commonly applied to this species was
Parker’s R. communis (Parker, 1914). Specimens of both R. anxia and R. querula were
determined under this name generally between 1914 and 1928. The revision provided by
Hall (1928) placed R. communis as a synonym of R. pallinervis. For several years, specimens
of R. anxia and R. querula were placed under this name. Aldrich (1930) examined types of
American Sarcophagidae in European museums and he provided the mistaken synonymy
of R. anxia and R. querula with R. lherminieri. Both R. anxia and R. querula were identified as
R. lherminieri until Dodge (1956) was able to provide characters to separate R. querula
from R. anxia. Therefore, references before 1956 concerning one of the earlier names must
be considered as potentially referring to R. anxia, unless otherwise indicated by the
author’s figures or comments.
Unfortunately, Dodge (1956) continued the improper usage of R. lherminieri as the
senior name of R. anxia. Discussions with Dr. W. L. Downes, Jr. (deceased) by one of us
21
(GAD) provided some insight into how this mistake occurred. Both Downes and Dodge
independently discovered characters to separate what was considered R. lherminieri into
two species. The type of R. lherminieri was not actually seen by either of these specialists,
but specimens representing the two species were independently sent to E. Seguy at the
National Museum of Natural History in Paris for comparison with the type. The best
comparison that Seguy could make indicated that R. lherminieri = R. anxia. Unfortunately,
neither specialist thought to include a third species, R. ochracea, which was originally
considered just a variation of R. communis by Aldrich. In 1986, one of us (GAD) was able to
examine the type of R. lherminieri at the Paris Museum and established from the holotype
that R. lherminieri = R. ochracea. During the same year GAD was able to examine the types
of Walker at the British Museum (Natural History). Direct examination of the types resulted
in the currently recognized synonymy given in Pape (1996).
Ravinia anxia shows a great deal of intra-specific variation in the shape of the male
and female genitalia. A close study of this species over its wide geographical range has led
to suspicion that this high amount of variation may indicate that this name may apply to
more than one species. Two larval forms are easily recognized from puparia pinned with
reared flies and this is mentioned by Dodge (1956). The first form, which is very common
in eastern North America, has very darkly pigmented dorsal tubercles. The second form
lacks these conspicuous tubercles. Many genitalic dissections of specimens from pivotal
areas, such as Arizona, Oregon, and Mexico, have led us to believe that R. anxia may
represent a species complex. However, no definitive morphological characters have been
found that will allow consistent separation of this species into smaller units. Genetic
studies, like the one reported here, in association with multiple rearings of this species
22
(especially from western localities) should be very useful for determining the range of
variation present in this species.
Morphological diagnoses for R. anxia, R. floridensis, R. lherminieri,
and R. querula
Ravinia anxia is easily separated from R. floridensis by its gray legs and palps. It can
be separated from R. lherminieri by the gray coloration of its head and abdomen. Males can
usually be separated from R. querula and R. lherminieri by the lesser developed lateral
scutellar setae, as compared to the subapical scutellar setae. The trapezoidal appearance of
sternite 7 of the female separates this species from the more rectangular sternite 7 of R.
querula. The lack of golden pruinosity on the genital sternites will easily separate females
from R. lherminieri.
Ravinia floridensis has been separated from other members of this group of species
based on its distinctive coloration. This species exhibits bright and conspicuous orange
palpi and legs.
Ravinia lherminieri is separated from the other species by the golden pruinose gena
and apical abdominal sternites, but with gray leg coloration.
Ravinia querula is separated from all the other species except R. anxia by its gray
pollinose gena and tergite 5. It can be separated from R. anxia in eastern North America by
the presence of two pairs of well-developed lateral scutellar setae adjacent to the preapical
scutellar setae. In western North America, the diagnostic shape of the male and female
genitalia are more useful in separating this species from R. anxia, as the size of the scutellar
setae are more variable in western individuals.
23
DNA extraction and amplification
DNA extractions using Qiagen’s DNeasy kit (Qiagen Cat. No. 69506) were performed
on a whole fly leg cut into thirds. A region of the mitochondrial genome encoding COI and
COII, spanning position 1460-3775 of the Drosophila yakuba Burla, 1954 (Clary and
Wolstenholme 1985; GenBank accession number NC_0013322), was amplified in fragments
using multiple primer pairs, listed in Stamper et al. (2013). PCR amplification solutions
consisted of: 200 μM dNTPs (Promega), 1μM of each primer, 1.5mM MgCl2, 0.25 units
Platinum Pfx DNA polymerase (Invitrogen), buffer (final concentration: 50mM Tris, pH 8.5,
4 mM MgCl2, 20 mM KCl, 500 μg/ml bovine serum albumin, 5% DMSO) (Idaho Technology),
and 3μl template DNA. Amplification was conducted using Rapid Cycler (Idaho
Technologies); thermocycling conditions are as follows: 95°C for 3min, followed by 35
cycles of 94 °C for 1s, 56-63°C for 1s, and 68°C for 20s. Forward strands only were
sequenced with the amplification primer using BigDye chemistry (PE Applied Biosystems)
on an automated ABI 3730 XL Instrument. Sequencing was performed by High Throughput
Genomics Unit at the Department of Genome Sciences, University of Washington.
Sequence assembly and characterization
Assembled sequences were edited in FinchTV v.1.4.0. (Geospiza, Inc.
www.geospiza.com) and aligned in Mesquite v.2.73 (Maddison and Maddison 2011). The
leucine tRNA gene separating COI and COII was excluded from analyses. The alignment
program Muscle v.3.8 (Edgar 2004) plug-in was used in Mesquite v.2.73 and the
subsequent alignment re-checked as both the translated amino acid sequence and the
nucleotide sequence. No indels were present among sequences, as such the final alignment
was identical to the preliminary Muscle v.3.8 alignment. We believe that nuclear insertions
24
of mitochondrial fragments (numts) were not erroneously sequenced due to the absence of
indels, absence of in-frame stop codons, taxon specific COI + COII primers used, and careful
examination of sequence characteristics such as amplicon length and quality score (phred)
of peaks (Song et al. 2008). Average Tamura-Nei corrected distances (Tamura and Nei
1993) within species-level groups and smallest Tamura-Nei corrected distances between
species-level groups were calculated in MEGA 5 (Tamura et al. 2011). The rate variation
among sites was modeled with a gamma distribution (shape parameter =1). The Tamura-
Nei + gamma corrected distance was selected as the best-fit model according to
PartitionFinder from among models implemented in the program MEGA 5. Smallest inter-
specific distances were used due to studies showing exaggeration of the divergence of
species sequence variability (i.e., barcoding gap; Meier et al. 2006).
Phylogenetic inference and model selection
To infer relationships, two methods were used in the phylogenetic analyses:
maximum likelihood (ML) using the program GARLI v2 (Genetic Algorithm for Rapid
Likelihood Inference; Zwickl 2006), and Bayesian inference (BI) using MrBayes v3.2.1
(Ronquist and Huelsenbeck 2003). Because both COI and COII make up a single linkage
group, the fragments are expected to share a single tree topology. However, there might be
differences in the evolutionary constraints across codon positions and possibly between
the two genes (Shapiro et al. 2006; Bofkin and Goldman 2007). To accommodate
gene/codon-position heterogeneity, DNA sequence data were partitioned into subsets, and
model parameters estimated across subsets.
PartitionFinder (Lanfear et al. 2012) was used to select the substitution models and
partitioning scheme for the ML and BI analyses. We compared 4 partition schemes:
25
unpartitioned; by gene (2 subsets); by codon position (3 subsets); and by both gene and
codon position (6 subsets). We calculated the log-likelihood scores under 56 models of
nucleotide evolution for each subset, including with and without gamma-distributed rates
(+G) and invariant sites (+I) for the ML analyses. For the Bayesian method, model selection
was limited to those that could be implemented in MrBayes, using the function “models =
mrbayes” in PartitionFinder. Models and partition schemes were selected using the Akaike
information criterion (AIC; Akaike 1973, 1974) test (with and without AICc correction).
For BI, we used the model + partitioning scheme chosen in PartitionFinder, and
sampled the posterior distribution by MCMC using MrBayes. Two independent Metropolis-
coupled Markov chain Monte Carlo (MC3) (Geyer 1991; Gilks and Roberts 1996) analyses
were run, each with one cold and three heated chains (temperature set to the default = 0.1)
for 1x106 generations and sampled every 500 generations; the first 25% of samples were
discarded as “burn-in.” Convergence on a stable posterior distribution was assessed by
using the program Tracer v1.5 (Rambaut and Drummond 2007) to visually examine the
long-term stability of the individual parameter estimates and the log-likelihood values, the
harmonic mean likelihood values of chains, the similarity of trees, branch lengths, and the
average standard deviation of split frequencies being less than 0.01 across runs. Posterior
probabilities (PP) were used to assess nodal support.
The model and partition schemes selected by PartitionFinder as best supported by
AIC (with and without AICc correction) were as follows: ML analysis: COI first: TIM + I + G;
COI second: F81 + I; COI third: TrN + G; COII first: TrN + I; COII second: HKY + I; COII third:
HKY + G; BI analyses: COI +COII first: GTR + I + G; COI + COII second: HKY + I; COI + COII
third: GTR + G.
26
The BI prior distributions were set as follows. 1st-position sites: substitution rates
across partition expressed as proportions of the overall rate sum, Dirichlet (1, 1, 1, 1, 1, 1);
substitution rates across site within partition, gamma distribution approximated with 4
categories with prior distribution for shape parameter uniform (0,200); proportion of
invariable sites uniformly distributed (0,1); nucleotide frequencies, Dirichlet (1, 1, 1, 1),
likelihood summarized over all rate categories in each gene. 2nd-position sites: transition
and transversion rates expressed as proportions of the rate sum with a (1,1) prior;
proportion of invariable sites, uniform (0, 1); nucleotide frequencies, Dirichlet (1, 1, 1, 1).
3rd-position sites: substitution rates across partition expressed as proportions of the
overall rate sum, Dirichlet (1, 1, 1, 1, 1, 1); substitution rates across site within partition,
gamma distribution approximated with 4 categories with prior distribution for shape
parameter uniform (0,200); nucleotide frequencies, Dirichlet (1, 1, 1, 1), likelihood
summarized over all rate categories in each gene.
ML analyses were conducted using the maximum likelihood function in Garli v2. The
optimal model + partition scheme was selected for the concatenated sequences with
PartitionFinder, using the AIC. Bootstrap resampling (Felsenstein 1985) was used to assess
nodal support with 1000 pseudoreplicates. Program defaults were used for stopping the
search. Maximum likelihood bootstrap probabilities (BPML) for the splits were mapped onto
the bootstrap consensus tree using SumTrees of the DendroPy v 3.12.0 (Sukumaran and
Holder 2010) program.
27
2.4 RESULTS
DNA sequence data were obtained for 106 individuals (including the 4 specimens
used as outgroups) comprising 10 of the 17 species of Ravinia that are found in North
America (Pape 1996; Pape and Dahlem 2010). The final molecular alignment consisted of
2239 base pairs (bp) of mtDNA sequence data: COI 1539 bp, with 350 parsimony-
informative characters and COII 700 bp, with 136 parsimony-informative characters. Data
matrices and trees used for this study have been uploaded to TreeBase
(http://purl.org/phylo/treebase/phylows/study/TB2:S15566). The phylogenetic
hypotheses inferred using both ML and BI analyses resulted in the same general topology,
with the BI consensus tree fully congruent with the best ML tree shown in Fig. 2.1. Average
Tamura-Nei corrected within-group pairwise (PW) distances and smallest Tamura-Nei
corrected between-group PW distances are shown in Table 2.2.
We find strong support for two clades within Ravinia that correspond to the groups
Ravinia and Chaetoravinia (Fig. 2.1; BPML=85; PP=1.0). This grouping is also supported by
the absence or presence of setae on the R1 wing vein in Ravinia and Chaetoravinia,
respectively (Lopes 1969). Within Chaetoravinia, our single specimen of R. errabunda
(Wulp, 1895) is sister to the other three species (Fig. 2.1; BPML=97; PP=1). Ravinia
stimulans (Walker, 1849), which is distinctive within Chaetoravinia based on male genital
morphology, showed very little intra-specific variation and was strongly supported as
sister to a clade comprising R. derelicta (Walker, 1853) + R. vagabunda (Wulp, 1895) (Fig.
2.1; BPML=99; PP=1.0). Our two specimens of R. vagabunda (from a single locality in New
Mexico) were monophyletic, but were weakly supported as nested within a paraphyletic R.
derelicta (Fig. 2.1; BPML=62; PP=0.87).
28
Within the Ravinia group, R. pusiola (Wulp, 1895) is sister to the other five species:
R. anxia and R. querula, R. floridensis and R. lherminieri, and Ravinia planifrons (Aldrich,
1916). Our 10 specimens of R. pusiola (all of which are from New Mexico) are monophyletic
and distinct from the other species (Fig. 2.1; BPML=99; PP=1.0). Within R. pusiola are two
distinct haplotype groups. Ravinia planifrons was distinct and strongly supported as
monophyletic (Fig. 2.1; BPML100; PP=1.0), and was sister to a clade consisting of R. anxia, R.
querula, R. floridensis, and R. lherminieri (Fig. 2.1; BPML=54; PP=0.65). Ravinia lherminieri
and R. floridensis formed a sister group to the R. anxia + R. querula group (Fig. 2.1; BPML=92;
PP=1.0). Our nine specimens of R. lherminieri sorted onto two distinct haplotype groups:
the R. lherminieri from Florida and Tennessee formed one highly supported clade (Fig. 2.1;
BPML=100; PP=1.0), which was sister to a paraphyletic group comprised of all of our R.
floridensis plus the R. lherminieri specimens from New York and West Virginia (Fig. 2.1;
BPML=100; PP=1.0). The R. lherminieri (FL+TN) haplotype group is distinct from the R.
lherminieri (NY+WV) haplotype group (smallest PW inter-group distance = 5.2%). The
average PW intra-group distances within each of these two R. lherminieri haplotype groups
are low (0.1% and 0.4%), and similar to the intra-group distances of other Ravinia species
(Table 2.2). Our six R. floridensis individuals had a haplotype that was nearly identical to
that of the R. lherminieri individuals collected from New York and West Virginia (smallest
PW inter-group distance = 0.0%), but distinct from that of the R. lherminieri collected from
Florida and Tennessee (smallest PW inter-group distance = 5.4%). In the group comprising
R. anxia and R. querula, we found highly supported paraphyletic relationships. Our 12
specimens of R. anxia and 16 specimens of R. querula sorted onto three distinct haplotype
groups. The R. anxia from Minnesota and New Mexico form a highly supported clade (Fig.
29
2.1; BPML=98; PP=1.0), that is sister to a smaller clade with two terminal nodes comprised
of: 1) R. querula from Oregon, Wisconsin, Georgia, and Virginia; and 2) R. anxia from
California and Oregon plus R. querula from California, Oregon, and New Mexico. Both of
these terminal nodes were highly supported in both the ML and BI analyses (Fig. 2.1;
BPML>95; PP=1.0). Over the large range of sampling for the species of both R. anxia and R.
querula, there are high levels of inter-, and intra-specific variation (Table 2.2).
2.5 DISCUSSION
Our analysis of the mtDNA phylogeny of Ravinia resulted in strong support for a
number of inter-specific relationships, while also suggesting that several species may be
either paraphyletic at the level of mtDNA genealogy or poorly delimited by current
morphological characters. At deeper phylogenetic levels, our data strongly support
separation of the species included in this study into two clades, which correspond to
previously recognized species groupings at the generic or subgeneric level (Ravinia and
Chaetoravinia). The separation of these two groups is further supported morphologically
based on the absence or presence of setae on the R1 wing vein. Together, the molecular and
morphological data suggest that Ravinia and Chaetoravinia are likely to be natural groups,
and merit recognition as subgenera of Ravinia or even as separate genera. However, formal
recognition of a change in taxonomic status should await inclusion of additional species,
particularly those expected to fall within the Adinoravinia group.
We sampled four species from the Chaetoravinia group: R. derelicta, R. errabunda, R.
stimulans, and R. vagabunda. Two of these species, R. derelicta and R. stimulans, are both
encountered very commonly and are both distributed throughout the continental United
States. Our collections include >40 specimens from widely distributed locations for each of
30
these two species. In contrast, R. errabunda (1 specimen in this study; Midwest and
Southern parts of the U.S. and Mexico) and R. vagabunda (2 specimens in this study;
Southwestern U.S. and Mexico) both have more restricted distributions and are both
encountered much less frequently. Ravinia derelicta, R. stimulans, and R. vagabunda form a
strongly supported clade (Fig. 2.1; BPML=99; PP=1.0), with R. errabunda as sister to that
group. Ravinia stimulans shows very limited mtDNA variation across the entire range
encompassed by our sampling and is sister to R. derelicta + R. vagabunda. Ravinia derelicta
may be paraphyletic with respect to mtDNA, because the two specimens of R. vagabunda
nested within R. derelicta. This is surprising, because the male genitalia of R. derelicta and
R. vagabunda are distinctly different from one another and these species are easily
separated by morphology. In terms of morphology, R. vagabunda and R. stimulans exhibit
much more similar male genitalia. However, it is not clear that this is a case of species-level
paraphyly because the branches that place R. vagabunda within, rather than as sister to, R.
derelicta are short, and the statistical support is weak (BPML=62; PP=0.87). Both of our
specimens of R. vagabunda came from a single locality in New Mexico, so additional
sampling would be important for fully understanding the relationship between these two
species. Regardless of the placement of R. vagabunda, it is clear that R. derelicta has more
genetic variability and intra-specific phylogenetic structure at the COI and COII loci,
compared to R. stimulans (Table 2.2).
Six species were sampled from the Ravinia group: R. anxia, R. floridensis, R.
lherminieri, R. querula, R. planifrons, and R. pusiola. Three of these species, R. anxia, R.
pusiola, and R. querula, are widely distributed and commonly encountered throughout the
continental United States and southern Canada. Ravinia lherminieri is mainly found in the
31
southeastern United States, while R. planifrons is more common in the Great Plains area of
the central United States. Ravinia floridensis has a distribution restricted to the states of
Florida and Georgia, but is commonly collected in these areas. Ravinia pusiola is strongly
supported as monophyletic (Fig. 2.1; BPML=99; PP= 1.0), and is strongly supported as the
first branching within the Ravinia group (Fig. 2.1; BPML=95; PP=1.0). Ravinia planifrons is
weakly supported as the next branching (Fig. 2.1; BPML=54; PP=0.65), with strong support
for its sister clade comprising R. anxia, R. floridensis, R. lherminieri, and R. querula (Fig. 2.1;
BPML=92; PP=1.0). In this clade, two groups were inferred to be monophyletic with strong
support, R. anxia + R. querula (Fig. 2.1; BPML=97; PP=1.0), and R. floridensis + R. lherminieri
(BPML=100; PP=1.0). Within each of these two latter clades, however, relationships are
more complex, with R. anxia, R. lherminieri, and R. querula all showing paraphyly in the
mtDNA tree (Fig. 2.1).
Ravinia floridensis and Ravinia lherminieri
Our results show that R. floridensis plus R. lherminieri form a strongly supported
mtDNA clade. Within this group are two distinct and strongly supported clades, but these
smaller clades do not correspond to the species as currently diagnosed by morphology. One
clade contains all of our R. floridensis specimens plus several R. lherminieri specimens from
New York and West Virginia, while the other clade comprises all remaining R. lherminieri
samples, including those captured within the geographic range of R. floridensis. The
discrepancy between the mtDNA tree and the current morphology-based species
identifications could have arisen in several ways. First, the current morphological species
delimitation could be correct, in which case the mtDNA genealogy would reflect either old,
retained polymorphism (incomplete lineage sorting; ILS) or post-speciation hybridization
32
followed by introgression of the R. floridensis mitochondrial genome into R. lherminieri. The
hybridization hypothesis seems the less likely of the two possibilities, due to the large
geographic distance between the flies that have R. floridensis-like mtDNA but were
morphologically identified as R. lherminieri and the known range of R. floridensis (Florida
and Georgia). Future data could be consistent with the hybridization/introgression
hypothesis if it turns out the R. lherminieri from localities that are geographically
intermediate between Florida and West Virginia, such as Georgia and the Carolinas, carry
the R. floridensis form of mtDNA while having nuclear genomes that are typical for R.
lherminieri. Second, R. floridensis and R. lherminieri could be a single species with both a
color polymorphism and a relatively deep intra-species divergence between two mtDNA
lineages. Finally, the deep division in the mtDNA phylogeny could reflect the actual species
boundary. This would have two implications: 1) the traits used to separate R. floridensis
and R. lherminieri under the current morphological definitions do not reflect species
boundaries (this would mean, in particular, that there are populations of R. floridensis that
lack the orange pedicel, palpi, and legs that are currently considered diagnostic for the
species); and 2) the true geographic range of R. floridensis extends much farther north than
is currently believed. Morphological variation within characters of both the male and
female genitalia in Ravinia lherminieri makes it difficult to use those traits to evaluate
species-level hypotheses. In our own collections it is difficult to find two male R. lherminieri
that exactly match one another in the fine structures of the aedeagus. Even collections from
a single locality on a single date showed morphological variability. Given this
morphological ambiguity, the most important future data for inferring species boundaries
in this clade might be expected to come from nuclear DNA.
33
Ravinia anxia and Ravinia querula
Ravinia anxia plus R. querula form another sister-species pair that shows divergent
mtDNA lineages that are discordant with current morphological species definitions. There
are 3 strongly supported mtDNA clades in this group: one comprised of only R. anxia
collected in two widely separated localities (Minnesota and New Mexico); one comprised of
only R. querula collected across a wide geographic area (e.g., Oregon, Wisconsin, Virginia);
and a third clade containing individuals assigned to both species according to the
morphological criteria. The mixed-species clade includes flies from the western U.S.
(California, Oregon, New Mexico). It is notable that our collections from Oregon include
members of both the R. querula-only and the mixed-species mtDNA lineages.
The morphological distinction between R. anxia and R. querula is well established,
so it is very unlikely that they would represent a single species. Otherwise, the same set of
possible explanations for discordance between mtDNA genealogy and morphological
species identifications apply to the R. anxia + R. querula clade: ILS, hybridization, or
incorrect morphological delimitation. There has been speculation in the past that R. anxia
may contain more than one species (see 2.3 Material and Methods: Nomenclatural Notes)
and the mitochondrial genealogy inferred in the present study (Fig. 2.1) would be
consistent with the presence of more than one species within what is now called R. anxia.
However, no adult traits have been identified that correspond to the two larval
morphotypes described by Dodge (1956) and we do not yet have specimens available with
which to test the hypothesis that the larval traits correspond to the division seen in the
mtDNA genealogy. As adults, R. anxia exhibit morphological variation over the wide
geographic range it inhabits, but more extensive molecular sampling will be required to
34
determine if molecular variation is associated with specific morphological characters that
could objectively identify smaller “species” units, or if R. anxia is a single, geographically
variable species. Mitochondrial data alone are likely not adequate for a full study of species
limits, so future studies of this potential complex should include multiple loci (both
mitochondrial and nuclear) and further morphological analysis of individuals (both larvae
and adults) sampled from across a larger geographic range.
Mitochondrial/Morphological Discordance and Molecular Species Identification
The goal of this study was to explore phylogenetic relationships within Ravinia, and
to that end all species identifications were performed using morphological characteristics.
Mitochondrial DNA data has been used for species identification in dipteran taxa (and
sarcophagid taxa in particular), dating back to Sperling et al. (1994). In recent years,
several studies have demonstrated the utility of mitochondrial data for species
identification in different dipteran taxa, using COI (Meiklejohn et al. 2011, 2012, 2013a;
Jordaens et al. 2013), COII (Guo et al. 2010), COI and COII (Tan et al. 2010), and sequence
data from other mtDNA genes and morphology (i.e., COI and CAD and morphology;
Meiklejohn et al. 2013b). Our results showing discordance between morphology and the
mtDNA phylogeny within the Ravinia group suggest that mtDNA might not be appropriate
for species identification purposes in at least some Ravinia unless future data support
delimitation of species along the lines suggested by the tree in Fig. 2.1.
2.6 CONCLUSIONS
This study is the first to analyze the phylogenetic relationships within Ravinia using
DNA sequence data. We find strong support for a number of interspecies relationships
within Ravinia. At the deepest level, the mtDNA phylogeny is consistent with a hypothesis
35
that the previously recognized subdivisions Ravinia and Chaetoravinia are natural groups.
If this result holds for future studies that include more species (especially those expected to
be part of an Adinoravinia lineage), then formal recognition of subgeneric or generic status
for these three groups would be warranted. Molecular phylogenetic studies are often
characterized by limited sampling, and this study is no exception. We have sampled only
those species of Ravinia found in North America, but there is an extensive diversity of
Ravinia in Central and South America, plus one species found only in the Old World. Even
within North America there were a number of species for which we obtained no specimens
and others for which our sampling was limited, due to the inherent rarity of several species
(c.f. Lim et al. 2012). Nevertheless, an important result of this study – the discordance
between morphological species criteria and the mitochondrial phylogeny of 1) R. anxia and
R. querula, and 2) R. floridensis and R. lherminieri – was not substantially affected by limited
availability of specimens. The present data cannot, however, distinguish between the
various hypotheses (incomplete lineage sorting, introgression, undescribed species) to
explain those discordant patterns. Further research focusing on increased geographic
sampling of the species that exhibit discordance, plus in-depth sampling of the nuclear
genome and morphological analysis of those same species will be needed in order to
address these species delimitation problems. Further examination of the larger picture of
Ravinia phylogeny will require increased species sampling, especially including the Old
World and Neotropical species of Ravinia.
36
ACKNOWLEDGMENTS
We thank Amanda J. Stein (New York), Tom Lyons (New York), Ed Quin (Minnesota),
Kristel Peters (Minnesota), Eric Lobner (Wisconsin), Donna Watkins (Florida) and Gary
Steck (Florida) for their help in obtaining collection permits in their respective state’s
parks. We thank Rudolf Meier and two anonymous reviews for comments that greatly
improved this manuscript.
This project was supported by Grant No. 2005-DA-BX-K102 awarded by the
National Institute of Justice (NIJ), Office of Justice Programs, US Department of Justice. The
opinions, findings, and conclusions or recommendations expressed in this publication are
those of the authors and do not necessarily reflect the views of the Department of Justice.
The authors recognize no conflicts of interest.
37
2.7 TABLES
Table 2.1. Specimen locality data and references for COI and COII sequences included in
this study. Localities are in the U.S.A. and are reported as county and state.
Species Locality Sex GenBank Accession COI/COII
Voucher/Citation
Blaesoxipha arizona Pape Grant, NM M JQ806809/JQ806788 AW32/Stamper et al. 2013 Blaesoxipha cessator (Aldrich) Grant, NM M JQ806810/JQ806789 AW27/Stamper et al. 2013 Oxysarcodexia cingarus (Aldrich) Schuyler, NY M JQ806816/JQ806795 AP68/Stamper et al. 2013 Oxysarcodexia ventricosa (Wulp) Hamilton, OH M GQ223321 E8/Stamper et al. 2013 Ravinia anxia (Walker) Siskiyou, CA M JQ807069/KJ586403 AE81 Ravinia anxia (Walker) Jefferson, OR F KJ586343/KJ586404 AG51 Ravinia anxia (Walker) Jefferson, OR F KJ586344/KJ586405 AH52 Ravinia anxia (Walker) Jefferson, OR M KJ586345/KJ586406 AJ18 Ravinia anxia (Walker) Jefferson, OR F KJ586346/KJ586407 AK25 Ravinia anxia (Walker) Kandiyohi, MN F KJ586347/KJ586408 AT44 Ravinia anxia (Walker) Kandiyohi, MN F JQ807068/KJ586409 AT45 Ravinia anxia (Walker) Kandiyohi, MN F KJ586348/KJ586410 AT46 Ravinia anxia (Walker) Kandiyohi, MN F KJ586349/KJ586411 AT47 Ravinia anxia (Walker) Kandiyohi, MN M KJ586350/KJ586412 AT49 Ravinia anxia (Walker) Kandiyohi, MN M JQ807070/KJ586413 AT50 Ravinia anxia (Walker) Grant, NM F KJ586351/KJ586414 AW48 Ravinia derelicta (Walker) Columbia, FL F KJ586352/KJ586415 AA10 Ravinia derelicta (Walker) Columbia, FL F KJ586353/KJ586416 AA13 Ravinia derelicta (Walker) Liberty, FL F KJ586354/KJ586417 AA25 Ravinia derelicta (Walker) Liberty, FL F KJ586355/KJ586418 AA26 Ravinia derelicta (Walker) Madison, FL F KJ586356/KJ586419 AA38 Ravinia derelicta (Walker) Walton, FL F KJ586357/KJ586420 AB05 Ravinia derelicta (Walker) Washington, FL F KJ586358/KJ586421 AB41 Ravinia derelicta (Walker) Highland, FL F KJ586359/KJ586422 AC75 Ravinia derelicta (Walker) Highland, FL F KJ586360/KJ586423 AC76 Ravinia derelicta (Walker) Highland, FL M KJ586361/KJ586424 AC81 Ravinia derelicta (Walker) Highland, FL M KJ586362/KJ586425 AD01 Ravinia derelicta (Walker) Highland, FL M KJ586363/KJ586426 AD02 Ravinia derelicta (Walker) Highland, FL M KJ586364/KJ586427 AD06 Ravinia derelicta (Walker) Highland, FL F KJ586365/KJ586428 AD07 Ravinia derelicta (Walker) Highland, FL F KJ586366/KJ586429 AD11 Ravinia derelicta (Walker) Martin, FL M KJ586367/KJ586430 AD40 Ravinia derelicta (Walker) Martin, FL M KJ586368/KJ586431 AD42 Ravinia derelicta (Walker) Glenn, CA F KJ586369/KJ586432 AE57 Ravinia derelicta (Walker) Jefferson, TN F KJ586370/KJ586433 AN13 Ravinia derelicta (Walker) Jefferson, TN M KJ586371/KJ586434 AN18 Ravinia derelicta (Walker) Hamilton, OH F JQ807072/KJ586435 AZ09 Ravinia derelicta (Walker) Bland, VA F JQ807073/KJ586436 AZ62 Ravinia derelicta (Walker) Hamilton, OH M JQ807075/KJ586437 BA29 Ravinia derelicta (Walker) Hamilton, OH M JQ806817/JQ806796 BA30/Stamper et al. 2013 Ravinia errabunda (Wulp) Glenn, CA M KJ586372/KJ586438 AE51 Ravinia floridensis (Aldrich) Duval, FL M JQ807076/KJ586439 AA01 Ravinia floridensis (Aldrich) Duval, FL M JQ807077/KJ586440 AA03 Ravinia floridensis (Aldrich) Duval, FL F KJ586373/KJ586441 AA04 Ravinia floridensis (Aldrich) Sarasota, FL M JQ807078/KJ586442 AA60 Ravinia floridensis (Aldrich) Washington, FL M JQ807079/KJ586443 AB15
38
Ravinia floridensis (Aldrich) Martin, FL F KJ586374/KJ586444 AD41 Ravinia lherminieri (Robineau-Desvoidy) Madison, FL M JQ807080/KJ586445 AA36 Ravinia lherminieri (Robineau-Desvoidy) Jefferson, TN M JQ807081/KJ586446 AN14 Ravinia lherminieri (Robineau-Desvoidy) Jefferson, TN M JQ807082/KJ586447 AN15 Ravinia lherminieri (Robineau-Desvoidy) Jefferson, TN M JQ806818/JQ806797 AN16/Stamper et al. 2013 Ravinia lherminieri (Robineau-Desvoidy) Jefferson, TN M JQ807083/KJ586448 AN17 Ravinia lherminieri (Robineau-Desvoidy) Saratoga, NY F JQ807099/KJ586449 AQ58 Ravinia lherminieri (Robineau-Desvoidy) Saratoga, NY F KJ586375/KJ586450 AR49 Ravinia lherminieri (Robineau-Desvoidy) Kanawha, WV F JQ807084/KJ586451 AZ44 Ravinia lherminieri (Robineau-Desvoidy) Kanawha, WV M JQ807085/KJ586452 AZ58 Ravinia planifrons (Aldrich) Siskiyou, CA M JQ807088/KJ586453 AF06 Ravinia planifrons (Aldrich) Jefferson, OR M KJ586376/KJ586454 AG34 Ravinia planifrons (Aldrich) Jefferson, OR M KJ586377/KJ586455 AG49 Ravinia planifrons (Aldrich) Jefferson, OR M JQ807089/KJ586456 AK08 Ravinia pusiola (Wulp) Grant, NM F JQ807094/KJ586457 AW64 Ravinia pusiola (Wulp) Grant, NM M JQ807095/KJ586458 AW73 Ravinia pusiola (Wulp) Grant, NM M JQ807093/KJ586459 AX61 Ravinia pusiola (Wulp) Grant, NM M JQ807096/KJ586460 AX62 Ravinia pusiola (Wulp) Grant, NM M JQ806819/JQ806798 AX63/Stamper et al. 2013 Ravinia pusiola (Wulp) Grant, NM F KJ586378/KJ586461 AY38 Ravinia pusiola (Wulp) Grant, NM M JQ807090/KJ586462 AY39 Ravinia pusiola (Wulp) Grant, NM M JQ807091/KJ586463 AZ05 Ravinia pusiola (Wulp) Grant, NM M JQ807092/KJ586464 BA18 Ravinia pusiola (Wulp) Grant, NM M JQ807098/KJ586465 BA19 Ravinia querula (Walker) Siskiyou, CA M KJ586379/KJ586466 AF03 Ravinia querula (Walker) Clackamas, OR M KJ586380/KJ586467 AF16 Ravinia querula (Walker) Jefferson, OR M KJ586381/KJ586468 AG28 Ravinia querula (Walker) Jefferson, OR M KJ586382/KJ586469 AG30 Ravinia querula (Walker) Jefferson, OR M KJ586383/KJ586470 AG31 Ravinia querula (Walker) Jefferson, OR M KJ586384/KJ586471 AG32 Ravinia querula (Walker) Jefferson, OR M KJ586385/KJ586472 AG33 Ravinia querula (Walker) Crook, OR M KJ586386/KJ586473 AH42 Ravinia querula (Walker) Crook, OR F KJ586387/KJ586474 AH51 Ravinia querula (Walker) Jefferson, OR M KJ586388/KJ586475 AJ19 Ravinia querula (Walker) Jefferson, OR F KJ586389/KJ586476 AJ23 Ravinia querula (Walker) Jefferson, OR M KJ586390/KJ586477 AK05 Ravinia querula (Walker) Rabun, GA M KJ586391/KJ586478 AL32 Ravinia querula (Walker) Marathon, WI F KJ586392/KJ586479 AS25 Ravinia querula (Walker) Bland, VA M JQ807101/KJ586480 AZ61 Ravinia querula (Walker) Grant, NM M JQ807102/KJ586481 BA16 Ravinia stimulans (Walker) Volusia, FL F KJ586393/KJ586482 AD60 Ravinia stimulans (Walker) Volusia, FL M KJ586394/KJ586483 AD64 Ravinia stimulans (Walker) Volusia, FL M KJ586395/KJ586484 AD65 Ravinia stimulans (Walker) Volusia, FL M KJ586396/KJ586485 AD66 Ravinia stimulans (Walker) Volusia, FL M KJ586397/KJ586486 AD67 Ravinia stimulans (Walker) Volusia, FL M KJ586398/KJ586487 AD68 Ravinia stimulans (Walker) Newberry, SC F KJ586399/KJ586488 AE11 Ravinia stimulans (Walker) Newberry, SC M KJ586400/KJ586489 AE12 Ravinia stimulans (Walker) Saratoga, NY M JQ807105/KJ586490 AR33 Ravinia stimulans (Walker) Kittson, MN F KJ586401/KJ586491 AU03 Ravinia stimulans (Walker) Hamilton, OH F JQ807109/KJ586492 AZ10 Ravinia stimulans (Walker) Hamilton, OH F JQ807103/KJ586493 AZ11 Ravinia stimulans (Walker) Kanawha, WV F JQ807110/KJ586494 AZ45 Ravinia stimulans (Walker) Kanawha, WV F JQ807111/KJ586495 AZ46 Ravinia stimulans (Walker) Bland, VA F JQ807112/KJ586496 AZ60
39
Ravinia stimulans (Walker) Hamilton, OH M JQ807108/KJ586497 BA27 Ravinia stimulans (Walker) Hocking, OH F KJ586402/KJ586498 D45 Ravinia stimulans (Walker) Hamilton, OH M GQ223320 E7/Stamper et al. 2013 Ravinia vagabunda (Wulp) Grant, NM M JQ807106/KJ586499 BA15 Ravinia vagabunda (Wulp) Grant, NM M JQ807107/KJ586500 BA17
40
Table 2.2. Tamura-Nei + G corrected distances within and between indicated groups.
Average Intra-Specific Distances, % Ravinia anxia (CA + MN + NM+ OR) 1.6 R. anxia (CA + OR) 0.1 R. anxia (MN + NM) 0.1 Ravinia derelicta 0.9 Ravinia floridensis 0.2 Ravinia lherminieri (FL + NY + TN +WV) 3.2 R. lherminieri (NY+WV) 0.1 R. lherminieri (FL + TN) 0.4 Ravinia planifrons 0.2 Ravinia pusiola 1.0 Ravinia querula 1.0 Ravinia stimulans 0.4 Ravinia vagabunda 0.1
Smallest Inter-Group Distances, % R. anxia (CA + OR) vs. R. querula (Para) 0.0 R. anxia (NM +MN) vs. R. querula (Para) 2.6 R. anxia: CA + OR vs. NM + MN 2.6 R. querula (Mono) vs. R. querula (Para) 2.4 R. floridensis vs. R. lherminieri (FL + TN) 5.4 R. floridensis vs. R. lherminieri (NY + WV) 0.0 R. lherminieri: NY + WV vs. FL + TN 5.2 R. derelicta vs. R. vagabunda 1.1
41
2.8 FIGURES
Figure 2.1. The inferred phylogenetic hypothesis of Ravinia species relationships of North
America, using maximum likelihood analysis of mitochondrial COI and COII genes.
Support values for nodes are given in the following order: bootstrap support (%)
from ML analysis using GARLI/Bayesian posterior probabilities using MrBayes.
43
2.9 REFERENCES
Akaike, H. (1973). Information theory and an extension of the maximum likelihood
principle. In ‘Second International Symposium on Information Theory’. (Eds. B. N.
Petrov and F. Caski.) pp. 267-281. (Akademiai Kiado: Budapest, Hungary).
Akaike, H. (1974). A new look at the statistical model identification. IEEE Transactions on
Automatic Control 19, 716-723.
Aldrich, J. M. (1916). Sarcophaga and allies in North America, Entomological Society of
America. (Murphy-Divins Co. Press: Lafayette, IN, USA.)
Aldrich, J. M. (1930). Notes on the types of American two-winged flies of the genus
Sarcophaga and a few related forms, described by the early authors. Proceedings of
the United States National Museum 78, 1-39.
Bofkin, L., and Goldman, N. (2007). Variation in evolutionary processes at different codon
positions. Molecular Biology and Evolution 24, 513-521.
Caterino, M. S., Cho, S., and Sperling, F. A. H. (2000). The Current State of Insect Molecular
Systematics: A Thriving Tower of Babel. Annual Review of Entomology 45, 1-54.
Catleka, B. C., and McAllister, B. F. (2004). A genealogical view of chromosomal evolution
and species delimitation in the Drosophila virilis species subgroup. Molecular
Phylogenetics and Evolution 33, 664-670.
Clary, D. O., and Wolstenholme, D. R. (1985). The mitochondrial DNA molecule of
Drosophila yakuba: nucleotide sequence, gene organization, and genetic code.
Journal of Molecular Evolution 22, 252-271.
Dodge, H. R. (1956). New North American Sarcophagidae, with some new synonymy
(Diptera). Annals of the Entomological Society of America 49, 182-190.
44
Edgar, R. C. (2004). MUSCLE: multiple sequence alignment with high accuracy and high
throughput. Nucleic Acids Research 32, 1792-1797.
Felsenstein, J. (1985). Phylogenies and the comparative method. American Naturalist 125,
1-15.
Geyer, C. J. (1991). Markov chain Monte Carlo maximum likelihood. In ‘Computing Science
and Statistics: Proceedings of the 23rd Symposium on the Interface’. (Ed. E. M.
Keramides.) pp.156-163. (Interface Foundation: Fairfax Station, USA).
Gilks, W. R., and Roberts, G. O. (1996). Strategies for improving MCMC. In ‘Markov chain
Monte Carlo in Practice’. (Eds. W. R. Gilks, S. Richardson, and D. J. Spiegelhalter.) pp.
89-114. (Chapman & Hall: London, UK).
Giroux, M., Pape, T., and Wheeler, T. (2010). Towards a phylogeny of the flesh flies (Diptera:
Sarcophagidae): Morphology and phylogenetic implications of the acrophallus in the
subfamily Sarcophaginae. Zoological Journal of the Linnaean Society 158, 740-778.
Guo, Y. D., Cai, J. F., Li, X., Xiong, F., Su, R. N., Chen, F. L., Liu, Q. L., Wang, X. H., Chang, Y. F.,
Zhong, M., Wang, X., and Wen, J. F. (2010). Identification of the forensically
important sarcophagid flies Boerttcherisca peregrina, Parasarcophaga albiceps and
Parasarcophaga dux (Diptera: Sarcophagidae) based on COII gene in China. Tropical
Biomedicine 27, 451-460.
Hall, D. G. (1928). Sarcophaga pallinervis and related species in the Americas. Annals of the
Entomological Society of America 21, 331-352.
Hebert, P. D. N., Penton, E. H., Burns, J. M., Janzen, D. H., and Hallwachs, W. (2004a). Ten
species in one: DNA barcoding reveals cryptic species in the neotropical skipper
45
butterfly Astraptes fulgerator. Proceedings of the National Academy of Sciences 101,
14812-14817.
Hebert, P. D. N., Stoeckle, M. Y., Zemlak, T. S., and Francis, C. M. (2004b). Identification of
birds through DNA barcodes. PLoS Biology 2, 1657-1663.
Jordaens, K., Sonet, G., Richet, R., Dupont, E., Braet, Y., and Desmyter, S. (2013).
Identification of forensically important Sarcophaga species (Diptera:
Sarcophagidae) using the mitochondrial COI gene. International Journal of Legal
Medicine 127, 491-504.
Kutty, S. J., Pape, T., Wiegmann, B. M., and Meier, R. (2010). Molecular phylogeny of the
Calyptratae (Diptera: Cyclorrhapha) with an emphasis on the superfamily
Oestroidae and the position of Mystacinobiidae and McAlpine’s fly. Systematic
Entomology 35, 614-635.
Lanfear, R., Calcott, B., Ho, S. Y. W., and Guindon, S. (2012). PartitionFinder: Combined
selection of partitioning schemes and substitution models for phylogenetic analyses.
Molecular Biology and Evolution 29, 1695-1701.
Lim, G. S., Balke, M., and Meier, R. (2012). Determining species boundaries in a world full of
rarity: singletons, species delimitation methods. Systematic Biology 61, 165-169.
Lopes, H. S. (1969). Family Sarcophagidae. In ‘A catalogue of the Diptera of the Americas
South of the United States, Vol. 103’. (Ed. N. Papavero.) pp. 1-88. (Departmento de
Zoologia, Secretaria da Agricultura: São Paulo, Brazil).
Lopes, H. S. (1982). The importance of the mandible and clypeal arch of the first instar
larvae in the classification of the Sarcophagidae (Diptera). Revista Brasileira de
Biologia 26, 293-326.
46
Maddison, W. P., and Maddison, D. R. (2011). ‘Mesquite 2.73: a modular system for
evolutionary analysis.’ Available at http://mesquiteproject.org.
Meier, R., Shiyang, K., Vaidya, G., and Ng, P. K. L. (2006). DNA Barcoding and Taxonomy in
Diptera: A Tale of High Intraspecific Variability and Low Identification Success.
Systematic Biology 55, 715-728.
Meiklejohn, K. A., Wallman, J. F., and Dowton, M. (2011). DNA-based identification of
forensically important Australian Sarcophagidae (Diptera). International Journal of
Legal Medicine 125, 27-32.
Meiklejohn, K. A., Wallman, J. F., Cameron, S. L., and Dowton, M. (2012). Comprehensive
evaluation of DNA barcoding for the molecular species identification of forensically
important Australian Sarcophagidae (Diptera). Invertebrate Systematics 26, 515-
525.
Meiklejohn, K. A., Wallman, J. F., and Dowton, M. (2013a). DNA Barcoding Identifies all
Immature Life Stages of a Forensically Important Flesh Fly (Diptera:
Sarcophagidae). Journal Forensic Sciences 58, 184-187.
Meiklejohn, K. A., Wallman, J. F., Pape, T., Cameron, S. L., and Dowton, M. (2013b). Utility of
COI, CAD and morphological data for resolving relationships within the genus
Sarcophaga (sensu lato) (Diptera: Sarcophagidae): A preliminary study. Molecular
Phylogenetics and Evolution 69, 133-141.
Moore, W. S. (1995). Inferring phylogenies from mtDNA variation: Mitochondrial gene trees
versus nuclear gene trees. Evolution 49, 718-726.
Pape, T. (1994). The world Blaesoxipha Loew, 1861 (Diptera: Sarcophagidae). Entomologica
Scandinavica Supplement 45, 1-247.
47
Pape, T. (1996). Genus Ravinia Robineau-Desvoidy. In ‘Catalogue of the Sarcophagidae of
the world (Insecta: Diptera)’. pp. 284-290. (Associated Publishers: Gainesville, FL,
USA.)
Pape, T., and Dahlem, G.A. (2010). Sarcophagidae (Flesh Flies). In ‘Manual of Central
American Diptera, Volume 2’. (Eds. Brown, B.V. et al.) pp. 1313-1335. (NRC Research
Press: Ottawa, Ontario, CAN.)
Parker, R. R. (1914). Sarcophagidae of New England: males of the genera Ravinia and
Boettcheria. Proceedings of the Boston Society of Natural History 35, 1-77.
Pickens, L. G. (1981). The life history and predatory efficiency of Ravinia lherminieri
(Diptera: Sarcophagidae) on the face fly (Diptera: Muscidae). Canadian Entomologist
113, 523-526.
Rambaut, A., and Drummond, A. J. (2007). ‘Tracer (Version 1.4).’ Available at
http://beast.bio.ed.ac.uk/TRACER.
Reid, N. M., and Carstens, B.C. (2012). Phylogenetic estimation error can decrease the
accuracy of species delimitation: a Bayesian implementation of the general mixed
Yule-coalescent model. BMC Evolutionary Biology 12, 196-206.
Robineau-Desvoidy, J. B. (1863). Histoire naturelle des Dipteres des environs de Paris. Vol.
2, pp. 920. (Paris, France).
Ronquist, F., and Huelsenbeck, J. P. (2003). MrBayes 3: Bayesian phylogenetic inference
under mixed models. Bioinformatics 19, 1572-1574.
Shapiro, B., Rambaut, A., and Drummond, A. J. (2006). Choosing appropriate substitution
models for phylogenetic analysis of protein-coding sequences. Molecular Biology and
Evolution 23, 7-9.
48
Smith, M. A., Wood, D. A., Janzen, D. H., Hallwacks, W., and Hebert, P. D. N. (2007). DNA
barcodes affirm that 16 species of apparently generalist tropical parasitoid flies
(Diptera, Tachinidae) are not all generalists. Proceedings of the National Academy of
Sciences 104, 4967-4972.
Song, H., Buhay, J. E., Whiting, M. F., and Crandall, K. A. (2008). Many species in one: DNA
barcoding overestimates the number of species when nuclear mitochondrial
pseudogenes are coamplified. Proceedings of the National Academy of Sciences 105,
13486-13491.
Sperling, F. A. H., Anderson, G. S., and Hickey, D. A. A. (1994). DNA-based approach to the
identification of insect species used for post-mortem interval estimation. Journal of
Forensic Sciences 39, 418-427.
Spillings, B. L., Brooke, B. D., Koekemoer, L. L., Chiphwanya, J., Coetzee, M., and Hunt, R. H.
(2009). A new species concealed by Anopheles funestus Giles, a major malaria vector
in Africa. American Journal of Tropical Medicine and Hygiene 81, 510-515.
Stamper, T., Dahlem, G. A., Cookman, C., and DeBry, R. W. (2013). Phylogenetic relationships
of flesh flies in the subfamily Sarcophaginae based on three mtDNA fragments
(Diptera: Sarcophagidae). Systematic Entomology 38, 35-44.
Sukumaran, J., and Holder, M. T. (2010). DendroPy: A Python library for phylogenetic
computing. Bioinformatics 26, 1569-1571.
Tamura, K., and Nei, M. (1993). Estimation of the number of nucleotide substitutions in the
control region of mitochondrial-DNA in humans and chimpanzees. Molecular Biology
and Evolution 10, 512-526.
49
Tamura, K., Peterson, D., Peterson, N., Stecher, G., Nei, M., and Kumar, S. (2011). MEGA5:
Molecular Evolution Genetics Analysis using Maximum Likelihood, Evolutionary
Distance, and Maximum Parsimony Methods. Molecular Biology and Evolution 28,
2731-2739.
Tan, S. H., Rizman-Idid, M., Mohd-Aris, E., Kurahashi, H. and Mohamed, Z. (2010). DNA-
based characterization and classification of forensically important flesh flies
(Diptera: Sarcophagidae) in Malaysia. Forensic Science International 199, 43-49.
Wells, J. D., and Stevens, J. R. (2009). Molecular Methods for Forensic Entomology. In
‘Forensic Entomology: The Utility of Arthropods in Legal Investigations, 2nd Ed’.
(Eds. J. H. Byrd, and J.L. Castner.) pp. 439-455. (CRC Press: London, UK.)
Wiens, J. J., and Penkrot, T. L. (2002). Delimiting species based on DNA and morphological
variation and discordant species limits in spiny lizards (Sceloporus). Systematic
Biology 51, 69-91.
Wiens, J. J., and Servedio, M. R. (2000). Species delimitation in systematics: Inferring
diagnostic difference between species. Proceedings of the Royal Society of London B
267, 631-636.
Zwikl, D. J. (2006). ‘Genetic algorithm approaches for the phylogenetic analysis of large
biological sequence datasets under the maximum likelihood criterion.’ PhD thesis.
(The University of Texas: USA.)
50
Chapter Three:
Coalescent DNA-based species delimitation within the flesh fly genus Ravinia
Robineau-Desvoidy, 1863 (Diptera: Sarcophagidae).1
1 This manuscript has been formatted for submission to Systematic Biology
51
3.1 ABSTRACT
Species and their constituent populations are the units upon which evolution acts.
Traditionally morphological characters are the first tools used in the identification of
species boundaries. However, the increase in availability and decrease in cost of obtaining
genetic data in non-model organisms has allowed researchers to go beyond the
conventional aspect of naming species to understanding the processes that occur during
speciation events. Additionally, using genetic data is helpful in situations where other lines
of evidence (e.g., morphology) are not available or are in disagreement. In this study we
investigate species limits in a complex of two morphologically variable flesh fly species
(Sarcophagidae: Ravinia) found in the United States. Multiple approaches were used in
inferring species boundaries; including multilocus coalescent-based species-tree methods,
population genetic analyses, migration analyses, and standard phylogenetic inference. We
find three morphologically cryptic lineages that are consistently recovered in analyses. To
prevent continued confusion within this complex we suggest the names Ravinia anxia
(sensu stricto), Ravinia n.sp., and Ravinia querula (sensu stricto) be used when referring to
these individuals until formal species descriptions can be conducted. Through this study
we clarify the species limits in this morphologically cryptic group and provide a backbone
for future species descriptions using morphological characters.
Keywords: delineation, cryptic species, speciation, multilocus coalescent, species-tree,
molecular, phylogenetics, North America
52
3.2 INTRODUCTION
Species, and their constituent populations, are the units upon which evolution acts.
The ability to accurately identify and delimit species boundaries is crucial for research, but
is also important for public policy (e.g., conservation: Lowenstein et al. 2009; disease
vectors: Spillings et al. 2009). Species delimitation can be described as the process through
which species boundaries are determined, new species are discovered, and currently
recognized species are found to be synonymous (Wiens 2007). Here we delimit members of
the flesh fly genus, Ravinia, using genetic data as the primary data for species delimitation
and provide evidence for recent (or ongoing) interspecific gene flow. To accomplish this we
utilize DNA-based species-tree and coalescent-based methods, phylogenetic inference,
population structure analyses, and migration analyses between identified species groups.
Despite considerable debate, a widely held view is that species are independently-
evolving metapopulation lineages, as described by the General Lineage Concept (GLC; de
Queiroz 2007). This concept has had a direct impact on our understanding of species, in
that multiple lines of evidence (i.e., reproductive isolation, monophyly, behavior,
morphology) can be used as secondary criteria to identify lineage divergence. Under the
GLC, conflict among alternative species criteria can be attributed to the variation of when in
time the lineage divergence, or speciation event, is estimated to have occurred. Recognizing
species as lineages provides a justification for the use of genetic data in the delimitation of
species, as genetic data reflect processes that were involved in the speciation event (i.e.,
lineage divergence). Speciation events and lineage divergences are recorded through the
sorting of ancestral alleles in the genome of an organism. The genetic record can then be
used to delimit species without the restrictive requirement of monophyletic loci (Knowles
53
and Carstens 2007). This sorting of ancestral alleles is the basis of coalescent theory
(Klingman 1982; Hudson 1991), which models species as lineages, and allows one to
estimate the probability of allele fixation within said lineages. Consequently, through the
application of the GLC, multilocus DNA sequence data are increasingly being used in
species delimitation studies.
The increased availability of multilocus data sets and the recent advent of analytical
methods capable of estimating migration rates have provided the ability to analyze species
both as independent lineages and as dynamic entities. While recent (or ongoing)
interspecific gene flow is a highly debated topic and can further complicate the recognition
of lineage divergence, several studies have provided examples of interspecific gene flow
occurring in a variety of taxa (e.g., Shaw 2002; Emelianov et al. 2004; Llopart et al. 2005;
Niemiller et al. 2008, 2012). A primary goal in species delimitation studies is discovering
characters (e.g., molecular, morphological, etc.) that accurately and uniquely describe a
species. However, migration and consequent gene flow from one population into another is
of great concern in species delimitations. Gene flow, whether a result of introgression,
migration, or hybridization, complicates the recognition of species boundaries by
increasing interspecific similarity. Gene flow may lead to genetic similarity among species
that are phenotypically divergent (Keck and Near 2010). Gene flow can also produce
phenotypic similarities (i.e., hybrids) among genetically divergent species depending upon
which genes recombine and which genes are sampled for analyses. This can be a significant
problem if working under the GLC; defining species as independently-evolving
metapopulation lineages (de Queiroz 2007). In fact, some of the most commonly used
coalescent-based methods incorporate an explicit assumption of no recent or ongoing
54
migration (e.g., *BEAST, Heled and Drummond 2010; Grummer et al. 2014; BP&P, Yang and
Rannala 2010). In addition, the effects caused by incomplete lineage sorting (ILS) are
difficult to separate from those effects resulting from gene flow across species boundaries
(Funk and Omland 2003). Although gene flow and ILS may produce a similar topology in
the inferred trees, these two processes have different implications in species delimitation
(Harrington and Near 2012). Taking a multilocus approach to species delimitation allows
for: 1) the correction of the stochastic nature of alleles through the coalescent caused by
ILS; 2) estimating the rates of gene flow between species units to determine if migration
has taken place.
The flesh fly family Sarcophagidae contains approximately 2600 described species
(Pape 1996), that can be found on every continent except Antarctica. A large number of
sarcophagid species are carrion-feeding insects, with some species’ life histories being
important to forensic entomology in estimations of time since death (post-mortem interval:
Greenberg 1991; Kutty et al. 2010). Many flesh fly species play an important role in the
recycling of nutrients in the environment. Typically, sarcophagid flies are identified to
species by an expert using morphological characters that are believed to be unique to that
group of organisms. A preponderance of the molecular studies on sarcophagid flies have
focused on using DNA data to provide a means to identify individuals to species, with
emphasis on the forensically important species (e.g., Meiklejohn et al. 2011, 2012, 2013a;
Jordaens et al. 2013). These studies have used either a single locus (COI, mitochondrial
DNA (mtDNA): Guo et al. 2010; Tan et al. 2010; Meiklejohn et al. 2011, 2012, 2013a;
Jordaens et al. 2013), or concatenated gene fragments (e.g., Meiklejohn et al. 2013b). The
data sets generated by these studies provide a framework within which genetic evidence
55
can be used to place an unknown specimen (a larva, for example) into a morphologically
defined species. Such studies presuppose that our current understanding of species
boundaries based on morphological traits studies are correct. In some Sarcophagidae,
however, suggestions have been made in the literature that there might be more species
than are currently recognized (e.g., Meiklejohn et al. 2011; Braga et al. 2013; Wong et al.
2015). In these studies, lineages that had been discovered by molecular methods could
then potentially be used as the backbone for identifying morphological traits useful in the
identification of the cryptic species (Meiklejohn et al. 2011; Braga et al. 2013). The
application of genetic data to questions of species limits has had success in a variety of
different taxa, where genetic data were used in both the discovery and validation of cryptic
species (e.g., Salter et al. 2013; Dumas et al. 2015). To date, studies of sarcophagid flies
have not used DNA-based species-tree approaches to species delimitation.
The genus Ravinia (Robineau-Desvoidy, 1863) is a common group of primarily New
World flesh flies, with 34 species recognized worldwide: 1 Palearctic species; 17 Nearctic
species; and 16 Neotropical species (Pape 1996; Pape and Dahlem 2010). Based on their
abundance in insect natural history collections, Ravinia are among the most commonly
collected flies of the family Sarcophagidae in North America (Wong et al. 2015).
Interestingly, unlike other sarcophagid genera, Ravinia larvae are primarily coprophagous.
A recent study found discordance between the mtDNA phylogeny and the
morphological characters used to identify some species within this genus (Wong et al.
2015). Two species within this genus, Ravinia anxia (Walker, 1849) and Ravinia querula
(Walker, 1849), can be found throughout much of continental North America in high
abundance. The taxonomic history of these two species is complex, with several taxonomic
56
revisions having occurred due to mistaken identifications by various authors (Wong et al.
2015). The mtDNA phylogeny given by Wong et al. (2015), is consistent with the
hypothesis that more than two species are present within the R. anxia – R. querula complex.
In order to test this hypothesis with multiple unlinked markers, we obtain DNA sequence
data from seven gene fragments. We then apply a combination of several species
delimitation and population genetic methods to test for lineage divergence at the species
level within R. anxia and R. querula. Furthermore, the presence of recent or ongoing gene
flow is estimated between species to determine if isolation with migration has occurred.
3.3 MATERIALS AND METHODS:
Specimen collection and sampling
Flies were collected using a net or by baited traps (using dog dung or chicken liver).
Specimens were frozen and taken to the lab for morphological identification, and molecular
analysis (based on Stamper et al. 2013). To preserve morphological structures important
for species identifications, DNA was extracted from 1-3 legs for use in analyses. Leg tissue
was transferred to a vial of 95% EtOH stored at -80C until DNA extraction. Specimens
were then pinned and are retained as vouchers (currently held by GAD, these will be placed
in an established natural history collection upon completion of current studies).
Specimens were collected from multiple localities throughout the continental United
States. At each locality we attempted to collect multiple individuals, although at some
localities this was not possible and only a single individual could be collected. Ravinia has a
broad distribution across North America, with some species (e.g., Ravinia anxia) found
throughout the continental United States, Canada, and northern parts of Mexico. Population
sampling of flying insects is also complicated by the large range that individuals can travel
57
over the course of their lifetime (e.g., Cochliomyia hominivorax; Mayer and Atzeni 1993). As
such, we attempted to collect individuals from different geographic localities for those
species that delimitation efforts would be focused on for this study: Ravinia anxia and R.
querula. Specimens in this study have been assigned a unique specimen number that can be
found in Table S3.2.
Based upon both morphological and molecular evidence (Lopes 1982; Pape 1994;
Giroux et al. 2010; Stamper et al. 2013; Wong et al. 2015), phylogenetic trees of R. anxia
and R. querula, were rooted using sequences from an outgroup taxon: Ravinia floridensis
(Aldrich, 1916).
DNA extraction and amplification
Genomic DNA extractions were performed on whole fly legs, cut into thirds, using
Qiagen’s DNeasy kit (Qiagen Cat. No. 69506). Seven separate gene fragments were PCR-
amplified, including two mtDNA and five nuDNA gene fragments (cytochrome oxidase
subunit I (COI), cytochrome oxidase subunit II (COII) mtDNA; elongation factor 1 alpha (EF-
1a), glucose-6-phosphate dehydrogenase (G6PD), phosphoenolpyruvate carboxykinase
(PEPCK), period (Per), white (Wt) nuDNA). PCR amplification solutions consisted of: 200
μM dNTPs (Promega), 1μM of each primer, 1.5mM MgCl2, 0.25 units Platinum Pfx DNA
polymerase (Invitrogen), buffer (final concentration: 50mM Tris, pH 8.5, 4 mM MgCl2, 20
mM KCl, 500 μg/ml bovine serum albumin, 5% DMSO) (Idaho Technology), and 3μl
template DNA. PCR thermocycling was conducted on Rapid Cycler (Idaho Technologies)
instruments. A region of the mitochondrial genome encoding COI and COII was amplified
using primer pairs with protocols listed in Stamper et al. (2013). Nested PCR was used to
amplify sequencing products for the nuclear genes: EF-1a, G6PD, PEPCK, Per, Wt. Primer
58
construction information and protocols for amplification (EF-1a, G6PD, PEPCK, Per, Wt) are
given in the Supplemental Materials (Appendix S3.1). Forward and reverse strands were
sequenced with the amplification primers using BigDye Terminator v.3.1, post reaction dye
terminator removal using Agencourt CleanSEQ and sequence delineation on an ABI prism
3730x/ with base calling and data compilation. Sequencing was performed by Beckman
Coulter Genomics (Danvers, MA).
Sequence Assembly and Characterization
Sequences were viewed in FinchTV v.1.4.0 (Geospiza, Inc. www.geospiza.com), and
assembled and edited in CLC Main Workbench v.7.0.3 (CLC Bio www.clcbio.com).
Assembled sequences were aligned in Mesquite v.2.73 (Maddison and Maddison 2011)
using the plug-in Muscle v.3.8 (Edgar 2004). The resulting alignment was re-checked as
both the translated amino acid sequence and the nucleotide sequence. No insertions or
deletions (indels) were present among sequences obtained for the genes (COI, COII, EF-1a,
G6PD, PEPCK, Per, Wt) as such the final alignment was identical to the preliminary Muscle
alignment. Based on the use of taxon specific primers, absence of in-frame stop codons, and
examination of sequence characteristics (i.e., amplicon length, quality score (phred) of
peaks; Song et al. 2008), we believe that nuclear insertions of mitochondrial fragments
(numts) and silenced genes were not erroneously sequenced.
Nuclear gene sequences were further analyzed using the program PHASE v.2.1
(Stephens et al. 2001; Stephens and Donnelly 2003) to resolve heterozygous haplotypes.
The web-based program SeqPHASE (Flot 2010) was used to convert DNA sequence data
into PHASE input format. Phased sites with 0.90 pp (posterior probability) and higher were
retained; heterozygous sites below this threshold use standard ambiguity codes. TOPALi
59
v.2.5 (Milne et al. 2009) was used to test for recombination with default program settings
using the Difference of Sums of Squares (DSS) method. The program DNasp v.5 (Librado
and Rozas 2009) was used to evaluate genetic variability of sequenced loci, calculating
number of unique alleles (k), segregating sites (S), parsimony informative sites (PI),
number of haplotypes (h), haplotype diversity (Hd) and nucleotide diversity (π).
Individual and Concatenated Gene-Tree Analyses
Models of DNA sequence evolution and gene partition were estimated with
PartitionFinder (Lanfear et al. 2012) using Akaike Information Criterion (AIC; Akaike 1973,
1974), with and without AICc correction. Due to restriction in model choice/partition,
model selection was limited to those that could be implemented in MrBayes (for Bayesian
inference) and RAxML (for maximum likelihood), using the function “models=mrbayes” or
“models=raxml”, respectively. To infer relationships, we analyzed mitochondrial and
nuclear loci with maximum likelihood (ML) using the program RAxML v.8.0.20 (Stamatakis
2014), and Bayesian inference (BI) using MrBayes v.3.2.1 (Ronquist and Huelsenbeck
2003).
For BI, two independent Metropolis-coupled Markov chain Monte Carlo (MC3)
(Geyer 1991; Gilks and Roberts 1996) analyses were used to sample the posterior
distribution using default settings in MrBayes. Each analysis had one cold and three heated
chains, sampling every 500 generations for 1x107 generations with the first 40% of
samples discarded as burn-in. Convergence on a stable posterior distribution was
determined using the program TRACER v.1.6. (Rambaut et al. 2014) to examine the
stability of parameter estimates, the harmonic mean likelihood values of chains, similarity
of trees, branch lengths, and average standard deviation of split frequencies being less than
60
0.01 between runs. Maximum likelihood analyses were conducted in RAxML v.8.0.20
(Stamatakis 2014). Due to restrictions in model choice when using RAxML, gene-trees and
the concatenated data set analyses were analyzed using subsets of the GTRGAMMA model
(Appendix S3.2). Nodal support was assessed with 1000 bootstrap replicates. The 70%
majority rule consensus tree of replicates was constructed using SumTrees of the
DendroPy v.3.12.0 (Sukumaran and Holder 2010) program.
Two concatenated data sets were used to perform the phylogenetic reconstruction.
We concatenated the phased nuclear loci (EF-1a, G6PD, PEPCK, Per, Wt) into a single data
set and analyzed these data separately before adding the mitochondrial data (COI, COII)
and repeating analyses. This was done to determine the influence that the mitochondrial
data have on the phylogenetic inferences. For the BI analyses of the concatenated data sets,
the number of generations was extended to 2x107, with the first 40% discarded as burn-in
to help with measuring convergence of runs. For the concatenated ML analyses, the
number of bootstrap replicates was increased to 5000. Individual gene-tree analyses were
inferred using the same parameters, except with the model + partitioning scheme selected
by PartitionFinder changed.
Coalescent-based Species Delimitation
A Bayesian implementation of the GMYC model (Pons et al. 2006), bGMYC (Reid and
Carstens 2012) was used as a species discovery method using the COI and COII data set.
This method, which does not require the a priori assignment of individuals to species
categories, identifies the species boundary through shifts in branching rates on a tree of
multiple species and populations (Monaghan et al. 2009). Since branching rates differ
between species and within species, the GMYC assesses the point of highest likelihood of
61
the transition between these two modes of lineage evolution (Pons et al. 2006; Fontaneto
et al. 2007). The independent lineages recovered from the GMYC analyses are the putative
species. The bGMYC is a Bayesian implementation of the GMYC model that samples the
posterior distribution of sampled gene-trees to incorporate gene-tree uncertainty (Reid
and Carstens 2012). BEAST v.1.8.0. (Drummond et al. 2012) was used to estimate
ultrametric gene-trees under a lognormal relaxed molecular clock using a MCMC run of
5x107 generations, sampling every 5000 generations with the first 40% of sampled trees
discarded as burn-in. The posterior distribution was trimmed to 100 trees, by evenly
sampling the posterior, and used as input for the bGMYC. Default settings were used;
starting number of species was set to half the total number of terminals. The analysis was
run on R. anxia and R. querula individuals + R. floridensis as outgroup. For all analyses,
bGMYC was run for 5 x 104 generations, with first 50% discarded as burn-in, and sampled
every 100 generations.
The program Bayesian Phylogenetics and Phylogeography (BP&P; Yang and Rannala
2010) uses a species validation-based approach implemented in a Bayesian coalescent
framework that is able to assign individuals to species using multilocus genetic sequence
data. BP&P calculates the posterior probabilities of species delimitation models using
reversible-jump Markov chain Monte Carlo (rjMCMC) algorithms. This method is able to
use the concordance of gene-trees across multiple loci but does not rely on reciprocal
monophyly as the determinant for multiple species (Zhang et al. 2011). Model assumptions
in BP&P include free recombination between loci, no gene flow or migration, neutral
evolution of loci, and no recombination within loci. BP&P also requires the use of a guide-
tree to decrease computation difficulty, and an accurate guide-tree is critical to the
62
accuracy of the species delimitation (Leaché and Fujita 2010). Therefore several species
hypotheses were generated that would be tested within BP&P to insure all possible species
would be represented in the guide-tree. First, morphologically identified species were
treated as species (R. anxia and R. querula). Second, the mtDNA gene-tree was used as a
guide for treating inferred terminal groups (Clade 1, 2, and 3; Fig. 3.1) as species. Third,
morphologically identified species and mtDNA gene-tree terminal groups were nested
together and treated as species. This third species hypothesis resulted in 4 hypothetical
species groups: Clade 1, Clade 2, Clade 3-R. querula, Clade 3-R. anxia.
Multiple analyses were run with varying priors ( and ) to determine how these
influence results, as suggested by Leaché and Fujita (2010). Parameters of ancestral
population size (ϴ) and root age (τo) were selected in order to mimic: large ancestral
populations and deep divergences (BP1); small population size and shallow divergences
(BP2); large populations with shallow divergence (BP3); small population size and deep
divergences (BP4). The BP&P analyses used the following settings: species delimitation
was set to 1, algorithm set to 0, and the fine tune parameter () set to 10. Analyses were
run for 5x105 generations, discarding the first 1x104 as burn-in, with a sampling interval of
5. Analyses were then repeated with starting seed change to check consistency between
runs. Strong support (pp>0.95) was required across both runs to retain a given node,
indicating lineage splitting.
The program *BEAST (Heled and Drummond 2010) implemented in the program
BEAST (Drummond et al. 2012) was used to estimate species-trees under the multispecies
coalescent model directly from the sequence data set containing the 2 mtDNA loci and 5
nuDNA loci. *BEAST was not originally developed for use as a species delimitation method
63
– but as a program meant for estimation of a species-tree under the multispecies coalescent
model. A method for species delimitation can be adopted, however, through the
incorporation of marginal likelihood estimation (MLE) to quantify the fit of each species
delimitation hypothesis (Grummer et al. 2014). Each *BEAST analysis requires the a priori
assignment of individuals to species categories to reconstruct the topology of the species-
tree. By conducting multiple *BEAST analyses, varying a priori species assignments
according to the species delimitation hypotheses, and quantifying the MLE for each
analysis, a best-fitting scenario can be determined. Therefore, species groups were defined
using the same species hypotheses generated for BP&P analyses. This resulted in a total of
three species delimitation hypotheses: a morphologically identified species scenario (R.
anxia and R. querula); a genetically identified species scenario (Clade 1, 2, and 3; Fig 3.1); a
morphologically identified species and genetically identified species scenario (Clade 1,
Clade 2, Clade 3-R. anxia, Clade 3-R. querula; Fig 3.1). Marginal likelihood estimation (MLE)
was performed on each analysis using path sampling (PS: Gelman and Meng 1998; Lartillot
and Phillippe 2006)/stepping-stone sampling (SS: Xie et al. 2011) as implemented in
*BEAST. The PS/SS analyses have been shown to provide a more accurate MLE than the
harmonic mean method for comparing different evolutionary models (Xie et al. 2011).
Analyses were conducted using both a nuclear only data set and a nuclear + mitochondrial
data set. The uncorrelated lognormal relaxed clock (Drummond et al. 2006) was assumed
and a Yule model prior used for divergence times in the species-tree. The MCMC algorithm
was run for 5x107 generations, sampling every 5000 generations with a 40% burn-in.
Convergence was assessed using TRACER v.1.6.0., checking for stationary in likelihood
64
scores and tree lengths. The model of nucleotide substitution was selected using
PartitionFinder, reducing selection to models that could be implemented in *BEAST.
Population Structure
The program STRUCTURE (Pritchard et al. 2000), which is a model-based, genetic
clustering method, was used to infer population structure from allele frequencies by
identifying clusters that are in Hardy-Weinberg and linkage equilibrium, without the a
priori assignment of population groups. STRUCTURE runs were conducted assuming
between 2 and 27 populations (K=2 through K=27), with each value of K replicated 5 times.
Analyses were conducted with a burn-in of 1x106 steps and 2x106 MCMC steps after burn-
in, used an admixture model, and allele frequencies were considered independent between
populations. The Evanno method (Evanno et al. 2005) implemented in STRUCTURE
HARVESTER (Earl and vonHoldt 2012) was used to estimate the optimal K value. The
program CLUMPP (Jakobsson and Rosenberg 2007) was used to summarize the data using
the “FullSearch” algorithm. The data were visualized using the program DISTRUCT
(Rosenberg 2004).
Migration & Isolation-with-migration models
To better understand the effect that migration could be having within this species
complex, migration analyses were performed using the programs Migrate-n v.3.6.4 (Beerli
and Palczewski 2010) and Isolation-with-Migration (IMa2; Hey 2010). Because both
coalescent-based analyses (*BEAST, bGMYC), population structure analyses (STRUCTURE),
and phylogenetic analyses (RAxML, MrBayes) found high support for the putative species
group made up of the individuals AW48 and AW70, these individuals were excluded from
downstream migration analyses. As such, these analyses included all individuals of R. anxia
65
and R. querula, excluding individuals AW70 and AW48. All data sets were analyzed using
the two-population assumption. In total, two data sets were used where the first data set
divided the putative species into morphologically identified R. querula + R. anxia, and the
second data set divided putative species into the individuals comprising Clade 2 + Clade 3
(Fig 3.1).
The program Migrate-n was used to estimate the long-term migration rates
(Mquerulaanxia, Manxiaquerula; MClade 2Clade 3, MClade 3Clade 2) and the effective population
sizes (anxia, querula; Clade 2, Clade 3) under a Bayesian inference model. Three independent
runs were performed to verify convergence, and uniform priors were used for all
parameters. Each run involved 1 long-chain, and four heated chains with 1.25x107 steps
following a burn-in of 2.5x104 steps, and a sampling increment of 25 steps. The four gene
flow models compared for the R. querula + R. anxia data set were: 1) Full 2-population gene
flow model; 2) R. querula to R. anxia migration; 3) R. anxia to R. querula migration; 4) Single
panmictic population. The four gene flow models compared for the Clade 2 + Clade 3 data
set were: 1) Full 2-population gene flow model; 2) Clade 2 to Clade 3 migration; 3) Clade 3
to Clade 2 migration; 4) Single panmictic population. Model choice was determined using
Bayes Factors and log marginal likelihood (Kass and Raftery 1995; Beerli and Palczewski
2010). Migrate-n is able to analyze gene flow among populations without the use of a
known tree topology, however it assumes that migration has been occurring between two
independent populations and cannot estimate divergence of populations through time.
The program IMa2 was used to estimate the migration rates (m1, m2), effective
population sizes (Nanxia, Nquerula, Nancestor), population divergence time (T), and the splitting
parameter (s) in the R. anxia/R. querula data set and the Clade 2/Clade 3 data set.
66
Preliminary runs were conducted to optimize the priors used in analyses. Final analyses
were run with an HKY mutation model (Hasegawa et al. 1985), a geometric heating scheme
(h1=0.96, h2=0.9), 15 chains, and a chain length of 2x107 with first 5 million steps
discarded as burn-in. To assess convergence, each run was repeated using the same prior
parameters with different random number seeds. The analyses were then repeated using
different starting priors for population size and migration rates. These tests were
compared to determine the similarity of results between runs. To ensure the proper mixing
of the Markov chain, the effective sample size (ESS) was monitored for each run.
3.4 RESULTS
Molecular Data
List of specimens, locality information, and GenBank accession numbers augmenting
those published in Wong et al. (2015) are provided in Table S3.2. Information for all loci
(unique alleles, segregating sites, nucleotide diversity) are found in Table 3.1. The final
concatenated molecular alignment consisted of 4905 base pairs (bp) of sequence data for 2
mtDNA (COI + COII) and 5 nuDNA (EF-1a, G6PD, PEPCK, Per, Wt) loci with the full data set
(R. anxia and R. querula + outgroup) including 31 individuals. TOPALi analyses indicated
possible recombination within the PEPCK nuDNA gene, but this is likely an artifact of very
small pairwise distances within regions of this gene where for some windows there is no
difference between sequences (McGuire et al. 1997).
Phylogenetic Reconstruction
The model of evolution and partitioning scheme selected as best by PartitionFinder
are given in the supplementary materials (Appendix S3.2). The BI and ML analyses, using
the mtDNA + nuDNA, and the nuDNA only data sets, resulted in the same general topology
67
and inferred a phylogeny congruent with previously inferred relationships using only
mtDNA (Fig 3.1; Wong et al. 2015). Ravinia anxia and R. querula are inferred as
paraphyletic sister species under current morphological-based assignments with high
nodal support (BPML=100; PP=1.0). The 31 individuals of R. anxia and R. querula sorted into
three distinct groups (Fig 3.1: Clade 1, Clade 2, & Clade 3). Ravinia anxia collected from
Minnesota and New Mexico form a highly supported clade (Fig 3.1, Clade 1: BPML=100;
PP=1.0), that is sister to a smaller clade with two terminal groups comprised of: 1) R.
querula from Oregon, Wisconsin, Georgia, and Virginia (Fig 3.1, Clade 2: BPML=100;
PP=1.0); 2) R. querula from California, Oregon, and New Mexico and R. anxia from
California and Oregon (Fig 3.1, Clade 3: BPML=100; PP=1.0). The inferred ML and BI
phylogenies obtained from single gene-trees showed varying degrees of resolution for R.
anxia and R. querula (Figs. S3.1-S3.7).
Test of Species Limits
Multiple approaches, including both discovery- and validation-based methods were
used to test species limits within the clades comprising all of the individuals assigned to R.
anxia and R. querula as currently defined by morphology. For the validation-based
methods, a priori partitions of individuals were generated based on competing hypotheses
of species delimitation scenarios. Species delimitation hypotheses were validated using
*BEAST and BP&P. For *BEAST, the marginal likelihood scores (with and without mtDNA)
estimated using PS/SS for the competing hypotheses are provided in Table 3.2 and the
inferred best species-tree is given in Fig S3.9. Results from *BEAST were congruent for
both the mitochondrial + nuclear data set and the nuclear-only data set (Table 3.2). There
was very strong support for the four-species delimitation model (DNA + Morphology Guide
68
Tree) between both data sets and PS/SS analyses (10< {2ln]>14.0, Kass and Rafferty 1995;
Table 3.2). However, nodal support for the best-fitting species-tree was weak for two out of
the three nodes (Fig. S3.9; PP<68). For BP&P, all guide trees regardless of ancestral
population sizes and divergence type priors returned equally supported, conflicting species
delimitation scenarios (Table S3.2).
For the discovery-based methods, bGMYC results using the mitochondrial data (COI
+ COII) detected more putative lineages than the other species delimitation methods (Fig.
S3.8). However, the majority of these putative lineages were found within a single clade
(Clade 2; Fig 3.1, Fig. S3.8) As a conservative measure to prevent over-splitting of potential
species groups, these lineages were merged resulting in a single putative species with a
posterior probability 0.5<PP<0.9 (Fig. 3.2). Both Clade 1 and Clade 2 were recovered as
highly supported putative species according to bGMYC posterior probabilities (Clade 1:
PP≥0.95, Clade 3≥0.90; Fig. 3.2).
Population Structure
The program STRUCTURE was used as a genetic clustering method that estimates
population structure based on Hardy-Weinberg and linkage equilibrium. The STRUCTURE
results estimated three species that are congruent with those obtained from the bGMYC
method (Fig. 3.2). The Evanno method (Evanno et al. 2005) was used to determine the best
partition, and a visual assessment of the STRUCTURE results corroborate the presence of
these three groups (Fig. 3.2). Due to the large geographic range that R. anxia and R. querula
inhabit, it is possible that isolation-by-distance may be a driving factor in the composition
of these three groups (Pritchard et al. 2000; 2010). However, given the large geographic
69
overlap of samples from Clades 2 and 3 (collected in California and Oregon), it is expected
that the effect of isolation-by-distance would have minimal impact.
Migration Analyses
Migration analyses using Migrate-n found high levels of uni-directional migration
between the inferred Ravinia anxia + Ravinia querula lineages and between the Clade 2 +
Clade 3 (Fig. 3.1) lineages. These levels of gene flow are suggestive that the BP&P analyses
could infer erroneous results in regards to delimitation models (Zhang et al. 2011).
Migrate-n results from the independent runs were all congruent, as such only results from
one run are shown (Table 3.3). For the R. anxia + R. querula data set, Model 2 (R. querula →
R. anxia migration) was selected as best according to the Bayes Factors with high
probability (P=0.997; Table 3.3). The number of migrants per generation based on 2
mtDNA loci and 5 nuDNA loci for R. querula → R. anxia was 1.02 individuals/generation
(95% highest posterior distribution; HPD = 0.08-3.17). For the Clade 2 + Clade 3 data set,
Model 2 (Clade 2 → Clade 3 migration) was selected as best according to the Bayes Factors
with high probability (P=0.997; Table 3.3). The number of migrants per generation for
Clade 2 → Clade 3 migration was 0.70 individuals/generation (95% highest posterior
distribution; HPD = 0.21-1.45).
Isolation-with-migration analyses using IMa2 also provided evidence of gene flow
between inferred R. anxia + R. querula lineages and between the Clade 2 + Clade 3 (Fig. 3.1)
lineages. All repeated runs produced similar results indicating significant migration events;
as such, only results from a single run are shown. ESS values for the time parameter ranged
from 74-29169 for the R. anxia + R. querula data set and 151-42102 for the Clade 2 + Clade
3 data set. Both data sets produced ESS values above the recommended cutoff value (Hey
70
2011). However, despite manipulation of prior parameters and extension of run times,
analyses did not converge on a stable ancestral population size. This is believed to be a
result of the data not containing enough information due possibly to a lack of depth in
phylogenetic signal for nuclear loci (Hey 2011). For the R. anxia + R. querula data set results
showed a statistically significant immigration event for both R. anxia → R. querula (M anxia >
querula; Table 3.4) and R. querula → R. anxia (M querula > anxia; Table 3.4). For the Clade 2 +
Clade 3 data set results showed a statically significant uni-directional immigration event
for Clade 2 → Clade 3 (MClade2 > Clade3; Table 3.4).
3.5 DISCUSSION
The main goal of this study was to test for evolutionary lineage divergence (i.e.,
speciation) at the species level within two currently recognized, Ravinia species. We find a
high level of genetic diversity and distinct lineage divergence within R. anxia and R. querula
that is not congruent with the current morphology-based taxonomy. Using DNA-based
coalescent species-tree methods, phylogenetic inference, population structure analyses,
and migration analyses, our result consistently indicate cryptic lineages within R. anxia and
R. querula, but the composition of these lineages varied across methods.
Comparison of coalescent-based species delimitation methods
Validation-based approaches are desirable because they can be computationally
tractable due to the a priori assignment of individuals to species. This can also be a
potential limitation of validation-based approaches because the accuracy of the species-
trees is constrained by the a priori hypothesized species groups. *BEAST analyses require
that individuals be partitioned to species groups before analyses. As such, several species
assignment models were generated from competing hypotheses based on groups obtained
71
from phylogenetic analyses (BI and ML), morphology-based assignments, or a combination
of the two. The best species delimitation model selected from this validation-based
approach recovered four evolutionary independent lineages (Fig. S3.9). This model splits
both morphologically identified species R. anxia and R. querula into each being composed of
two cryptic species. The lineage split that is seen in Clade 3 (Fig. S3.9) is between
individuals that for the most part occur in sympatry in California and Oregon and lack any
obvious geographic barriers. Leaché et al. (2014) conducted a recent simulation study
using *BEAST to look at how gene flow influences the topology and support of the species-
tree. This simulation study showed that paraphyletic gene flow, regardless of whether it
occurred at shallow or deep levels in the species-tree, can lead to strong support for the
incorrect topology and a sharp decline in posterior probability of clades (Leaché et al.
2014). The 4-species model, selected as best-fit by the PS/SS analyses implemented in
*BEAST, would have intermediate levels of paraphyletic gene flow and as such potential to
infer the incorrect species-tree with high support (Leaché et al. 2014). What is evident
from the *BEAST results is that when comparing the 4-species scenario and 3-species
scenario there is a relatively small difference between likelihoods, but a rather large
difference is seen between the 3-species scenario and the 2-species scenario (Table 3.2).
This suggests that while *BEAST may have difficulty choosing between a 4-species or 3-
species scenario, the elimination of the 2-species scenario as most likely appears to be
probable.
Approaches that are capable of delimiting lineages without the requirement of a
priori species assignments can be critical when there is a lack of evidence for pre-existing
divisions. Two discovery-based methods were used to identify the presence of evolutionary
72
lineage divergence: bGMYC and BP&P. The inferred lineages recovered using the multilocus
Bayesian method BP&P gave ambiguous results. Regardless of the prior parameters and
guide trees used, all species delimitations scenarios were equally and highly supported for
each model (PP=1.0; Table S3.2). An explicit assumption of BP&P is an absence of gene flow
between lineages (Yang and Rannala 2010). At low rates of migration (~0.1
individuals/generation), there is virtually no impact on BP&P analyses (Zhang et al. 2011).
At high migration rates (i.e., ≥10 individuals/generation), BP&P tends to infer only a single
species (Zhang et al. 2011). In contrast, at intermediate levels of migration (1.0-10
individuals/generation) BP&P is found to be indecisive (Zhang et al. 2011). Due to the fact
that intermediate levels of migration (0.7-1.02 individuals/generation) were estimated
within R. anxia – R. querula, BP&P is not a suitable multilocus species delimitation analysis
for this study.
The second discovery-based species delimitation method that we used was bGMYC.
The bGMYC results showed multiple potential coalescent units, particularly present among
Clade 2 individuals (Fig. 3.2, Fig. S3.8). Species delimitation is complicated within Ravinia,
and in systems that are genetically and geographically fragmented in general, with the
potential to oversplit species. While the bGMYC model has been used previously to delimit
insect species (Pons et al. 2011; Keith and Hedin 2012), it is most reliable among species
that exhibit a measurable difference in DNA similarity (i.e., barcoding-gap; Pons et al. 2006;
Satler et al. 2013). The bGMYC method identifies the species boundaries from the change in
branching rates in the gene-tree between the levels of species and populations (Pons et al.
2006). The branching rates are expected to differ at the speciation event, where lengths
between species are a result of speciation and extinction events (Nee et al. 1994), and
73
interspecific branch lengths a results of the coalescent (Rosenberg and Nordborg 2002;
Wakely 2006). However, coalescent events in the gene-tree are expected to be found within
clades where genetic structure is lacking because it is assumed that the individuals in these
clades are panmictic (Pons et al. 2006; Satler et al. 2013). The high amount of structure that
is seen within Clade 2 (Fig. 3.2) suggests that the demarcation inferred by the GMYC model
may be biased toward the present, and thus many coalescent units (i.e., species) would be
recognized. This problem in GMYC analyses has been recognized to exist in other taxa also
exhibiting conspicuous structure (Lohse 2009; Hamilton et al. 2011; Satler et al. 2013). We
suggest that this is what is occurring among individuals that make up Clade 2 (Fig. 3.2, Fig.
S3.8) and is likely the result of the large geographic ranges for which these particular
individuals were sampled. If the multiple coalescent units within Clade 2 (Fig. 3.2, Fig. S3.8)
are collapsed, they form a single unit with a relatively high nodal support (0.5< PP >0.9). In
addition, the merging of these coalescent units in Clade 2 would result in a species
delimitation scenario congruent with the population structure analyses, the lineages
recovered from concatenated ML and BI analyses, and the migration analyses.
Population Structure and Migration Analyses
As an alternative to coalescent-based analyses, we used the program STRUCTURE, a
genetic clustering algorithm that is able to incorporate gene flow between groups, utilize
multilocus data, and does not require a priori assignments. We wanted to further test the
lineages (Clades 1, 2, and 3; Fig. 3.2) recovered from the ML and BI analyses and see if
corresponding groups would be inferred from the STRUCTURE results. STRUCTURE
analyses resulted in three groups that corroborate with the Clades 1, 2, and 3 obtained
74
from ML and BI analyses (Fig. 3.2). A visual examination of the obtained STRUCTURE plot
(Fig. 3.2) shows little overlap between inferred groups.
Migration analyses (Migrate-n and IMa2) yielded results showing migration
between hypothetical species units regardless of the data set used. For the Migrate-n
analyses, the migration rate between putative species and the highest-posterior
distribution was decreased when the Clade 2 + Clade 3 data set was used, relative to
analyses using the morphologically identified R. querula + R. anxia data set (Table 3.4).
While migration between the putative species comprising Clade 2 + Clade 3 is still present
(0.7 individuals/generation), these results indicate there is likely to be more barriers to
gene flow than the species delimitation scenario present in the current taxonomy (R. anxia
+ R. querula; 1.02 individuals/generation). Isolation-with-migration analyses using IMa2
showed comparable levels of migration to the Migrate-n analyses. In the IMa2 analyses, the
migration rate between putative species was decreased with the Clade 2 + Clade 3 data set,
the migration was uni-directional, and the High/Low 95% HPD also decreased relative to
the R. anxia + R. querula data set (Table 3.4).
In both migration analyses, two species delimitation scenarios were considered. The
first, using the R. anxia + R. querula data set, assumes that those individuals
morphologically identified are the putative species. The second, using the Clade 2 + Clade 3
data set, assumes that those individuals defined by the Clade 2 lineage and Clade 3 lineage
are the putative species. In order to accept this first species delimitation scenario, one
would also have to accept a large migration rate between the putatively defined species.
The scenario that delimits the species into Clade 2 and Clade 3 putative species has much
lower levels of migration. It is expected that the more likely delimitation scenario would
75
have lower levels of migration between putative species (i.e., more barriers to gene flow)
(Mayr 1942, 1963). In addition, Wright (1931) showed that species cohesion breaks down
when gene flow is reduced below one migrant per generation. Thus, the migration rate
between the morphologically identified R. anxia and R. querula, according to Wright (1931),
would be sufficiently high enough to prevent lineage divergence.
3.6 CONCLUSIONS
The criteria that are used in determining if a population is a species, is an ongoing,
controversial topic. The use of genetic data for species discovery and delimitation has been
proposed as an approach that is objective and uses methods that are reproducible (Fujita
and Leaché 2011). There is debate though on the primary use of genetic data in species
delimitation studies and its ultimate utility (Bauer et al. 2011). However, it becomes
obvious in cryptic lineages with high levels of genetic diversity that, at least on some level,
accuracy in species delimitation will be a product of the data available and decisions that
are made by the researchers (Satler et al. 2013). This is a particular challenge when there is
incongruence among methods; where the researcher is left the task of choosing the most
appropriate species delimitation scenario. As such, it is important to employ a wide range
of methods to the data and, in short, select those scenarios that are most congruent across
methods (Carstens et al. 2013). Additionally, it is important to take a conservative
approach to species delimitation studies to prevent the oversplitting of species (Carstens et
al. 2013).
Across all methods, the species delimitation scenario that uses the current
morphological taxonomy, R. anxia and R. querula, is the least supported. This is particularly
evident in *BEAST analyses where the morphologically-based species scenario is separated
76
from the next most likely scenario by well over a hundred Bayes factors regardless of the
data set used (Table 3.2). The results obtained from bGMYC suggest that this particular
method, which uses a single locus, may not be suitable in the species delimitation of
Ravinia, where high levels of genetic diversity are present. However, a conservative
approach can be taken, where the multiple coalescent units of Clade 2 (Fig. 3.2, Fig. S3.8)
are merged, and obtain the 3-species delimitation scenario (0.5<PP<0.9; Fig. S3.8). Using
this conservative bGMYC approach would produce results congruent with those from the
migration analyses, phylogenetic inferences, and population structure analyses. In short
the migration analyses (Migrate-n, IMa2), population structure analyses (STRUCTURE),
and phylogenetic inference methods support the 3-species delimitation scenario. In
addition, it is clear that the *BEAST and bGMYC results do not support a 2-species
delimitation scenario based on the current morphological taxonomy. The incongruence
between these different methods (*BEAST and bGMYC vs. Migrate-n, IMa2, STRUCUTRE,
MrBayes, RAxML) is taken to be evidence that important methodological assumptions (i.e.,
migration/gene flow) have been violated (Carstens et al. 2013).
Given the strong support of the phylogenetic, population genetic, and migration-
with-isolation analyses we feel that these three species are distinct. This distinction,
however, is currently only found within the molecular data. Future additional evidence
obtained using other techniques (e.g., morphology) would strengthen the status of these
species, and additional efforts will be dedicated to such research. At this time, however, the
authors defer to the current conventions adopted in the International Code of Zoological
Nomenclature concerning the naming of new species (Articles 11-20; 1999). As such, to
prevent the use of nomina nuda the taxonomic rank and names will remain until further
77
research provides morphological descriptions of these new species. However, to decrease
confusion when referencing these three species in subsequent research, we suggest these
conventions be used when referring to these genetically delimited species: Ravinia anxia
(sensu stricto) = individuals comprising Clade 1; Ravinia querula (sensu stricto) =
individuals comprising Clade 2; Ravinia n.sp. = individuals comprising Clade 3.
78
ACKNOWLEDGMENTS
We thank Amanda J. Stein (New York), Tom Lyons (New York), Ed Quin (Minnesota),
Kristel Peters (Minnesota), Eric Lobner (Wisconsin), Donna Watkins (Florida) and Gary
Steck (Florida) for their help in obtaining collection permits in their respective state’s
parks.
This project was supported by Grant No. 2005-DA-BX-K102 awarded by the
National Institute of Justice (NIJ), Office of Justice Programs, US Department of Justice. This
project was supported by the Department of Biological Sciences and The Graduate Student
Governance Association of The University of Cincinnati. The opinions, findings, and
conclusions or recommendations expressed in this publication are those of the authors and
do not necessarily reflect the views of the Department of Justice or The University of
Cincinnati. The authors recognize no conflicts of interest.
79
3.7 TABLES
Table 3.1. Gene information for multilocus analyses. Includes name of locus, length (bp),
number of sequences (# Seq.), number of segregating sites (S), number of
parsimony informative sites (PI), nucleotide diversity (), number of haplotypes (h),
haplotype diversity (Hd), and model of sequence evolution for the R. anxia + R.
querula data set.
Locus bp # Seq. S PI h Hd Model COI/COII mtDNA 2241 29 31 31 0.01508 6 0.707 GTR + I + G EF-1a nuDNA 712 58 18 6 0.00307 10 0.699 SYM + I White nuDNA 589 60 34 28 0.01427 39 0.973 GTR + I + G Period nuDNA 512 58 22 17 0.00873 19 0.852 HKY + G G6PD nuDNA 485 60 92 87 0.01674 29 0.892 TrN + I + G PEPCK nuDNA 400 60 13 11 0.01174 6 0.615 SYM + G
80
Table 3.2. *BEAST marginal likelihoods. Estimated using mitochondrial and nuclear data
for each species delimitation model tested.
Mitochondrial and Nuclear Data Path-Sampling Stepping-Stone
Delimitation Model ln(Marginal Likelihood)
2ln(Bayes Factor)
ln(Marginal Likelihood)
2ln(Bayes Factor)
DNA + Morphology Guide Tree -10544.1* - -10544.7* Morphology Guide Tree -10724.5 180.4++ -10724.7 180.0++
Mitochondrial Guide Tree -10558.4 14.3++ -10558.7 14.0++
Nuclear Data
Delimitation Model ln(Marginal Likelihood)
2ln(Bayes Factor)
ln(Marginal Likelihood)
2ln(Bayes Factor)
DNA + Morphology Guide Tree -6362.0* - -6362.3* Morphology Guide Tree -6511.4 149.4++ -6511.9 149.6++
Mitochondrial Guide Tree -6383.5 21.5++ -6383.8 21.5++
Note: Marginal Likelihoods estimated using the Path-Sampling and Stepping-Stone Methods implemented in *BEAST *best-fitting hypothesis, under comparison with best-fitting hypothesis and alternative species delimitation hypothesis ++very strong support for best-fitting hypothesis (10< 2ln[BF]; Kass and Raftery 1995)
81
Table 3.3. Migration results using Migrate-n
Migration excluding Clade 1 (AW70 + AW48) individuals R. anxia/R. querula
data set Bezier
lmL Harmonic
lmL LBF
(Bezier) Choice
(Bezier) Model
Probability Full Model -9260.58 -8920.46 30.36 3 2.550*10-7
M querula -> anxia -9245.40 -8913.86 0.00 1 (BEST) 0.997 M anxia -> querula -9251.44 -8949.16 12.08 2 2.376*10-3
Panmictic -9314.11 -8913.53 137.42 4 1.44*10-30
Clade 2/Clade 3 data set
Full Model -9226.16 -8913.39 45.08 3 2.549*10-7
M Clade 2 -> Clade 3 -9203.62 -8933.50 0.00 1 (BEST) 0.997 M Clade 3 -> Clade 2 -9217.54 -8921.34 27.84 2 2.376*10-3
Panmictic -9313.68 -8927.61 220.12 4 1.441*10-30
82
Table 3.4. Mean estimates of population parameters from IMa2 analyses. Highest (H) and
Lowest (L) 95% HPD (highest posterior density) are given. Note: “M anxia > querula”
read as: immigration into R. anxia; “MClade2 > Clade3” read as: immigration into Clade 2.
†Indicates significant LLR test >0 (Nielsen and Wakeley 2001).
R. anxia + R. querula Data set Parameter Mean L(95%) H(95%) ϴanxia 1.032 0.448 1.683 ϴquerula 4.315 3.458 4.997 ϴancestral 4.549 3.732 4.997 M anxia > querula 0.310† 0.000 0.867 M querula > anxia 0.497† 0.165 0.881
Clade 2 + Clade 3 Data set Parameter Mean L(95%) H(95%) ϴClade 2 1.632 0.933 2.398 ϴClade 3 3.586 2.518 4.773 ϴancestral 4.488 3.583 4.997 MClade2 > Clade3 0.113† 0.008 0.248 M Clade3 > Clade2 0.027 0.248 0.080
83
3.8 FIGURES
Figure 3.1. Maximum likelihood tree, inferred using RAxML, of the concatenated (2mtDNA
loci + 5nuDNA loci) data set for all samples. Bayesian analyses, using MrBayes,
returned the same topology. The same topology was seen with the reduced data set
(with and without inclusion of mtDNA loci). Support values for nodes are ML-
bootstrap support (%)/BI-posterior probabilities.
85
Figure 3.2. Results from species delimitation and STRUCTURE analyses, represented on
BEAST tree from bGMYC of all sampling localities.
87
3.9 REFERENCES
Akaike H. 1973. Information theory and an extension of the maximum likelihood principle.
In: Petrov B.N., Caski, F., editors. Second International Symposium on Information
Theory. Budapest: Akademiai. p. 267-281.
Akaike H. 1974. A new look at the statistical model identification. IEEE T. Automat. Contr.
19:716-723.
Bauer A.M., Parham J.F., Brown R.M., Stuart B.L., Grismer L., Papenfuss T.J., Böhme W.,
Savage J.M., Carranza S., Grismer J.L., Wagner P., Schmitz A., Ananjeva N.B., Inger R.F.
2011. Availability of new Bayesian-delimited gecko names and the importance of
character-based species descriptions. Proc. Royal Soc. B. 278:490-492.
Beerli P., Palczewski M. 2010. Unified framework to evaluate panmixia and migration
direction among multiple sampling locations. Genetics. 185:313-326.
Braga M.V., Pinto Z.T., de Carvalho Queiroz M.M., Matsumoto N., Blomquist G.J. 2013.
Cuticular hydrocarbons as a tool for the identification of insect species: Puparial
cases from Sarcophagidae. Acta Trop. 128:479-485.
Carstens B.C., Pelletier T.A., Reid N.M., Satler J.D. 2013. How to fail at species delimitation.
Mol. Ecol. 22:4369-4383.
de Queiroz K. 2007. Species concepts and species delimitation. Syst. Biol. 56:879-886.
Drummond A.J., Ho S.Y.W., Phillips M.J., Rambaut A. 2006. Relaxed phylogenetics and dating
with confidence. PLoS Biol. 4:e88.
Drummond A.J., Suchard M.A., Xie D., Rambaut A. 2012. Bayesian phylogenetics with BEAUti
and BEAST 1.7 Mol. Biol. Evol. 29:1969-1973.
88
Dumas P, Barbut J., Ru B.L., Silvain J.-F., Clamens A.L., d’Alençon E., Kergoat, G.J. 2015.
Phylogenetic molecular species delimitations unravel potential new species in the
pest genus Spodoptera Guenée, 1852 (Lepidoptera, Noctuidae). PLoS ONE. 10(4):
e0122407.
Earl D.A., vonHoldt B.M. 2012. STRUCTURE HARVESTER: a website and program for
visualizing STRUCTURE output and implementing the Evanno method. Conserv.
Genet. Resour. 4:359-361.
Edgar R.C. 2004. MUSCLE: multiple sequence alignment with high accuracy and high
throughput. Nucleic Acids Res. 32:1792-1797.
Emelianov I., Marec F., Mallet J. 2004. Genomic evidence for divergence with gene flow in
host races of the larch budmoth. Proc. Royal Soc. B. 271:97-105.
Evanno G., Regnaut S., Goudet J. 2005. Detecting the number of clusters of individuals using
the software STRUCTURE: a simulation study. Mol. Ecol. 14:2611-2620.
Flot J.-F. 2010. SEQPHASE: a web tool for interconverting PHASE input/output files and
FASTA sequence alignments. Mol. Ecol. Resour. 10:162-166.
Fontaneto D., Herniou E.A., Boschetti C., Capriolo M., Melone G., Ricci C., Barraclough T.G.
2007. Independently Evolving Species in Asexual Bdelloid Rotifers. PLoS Biol. 5:e87.
Fujita M.K., Leaché A.D. 2011. A coalescent perspective on delimiting and naming species: a
reply to Bauer et al. Proc. Royal Soc. B. 278:493-495.
Funk D.J., Omland K.E. 2003. Species-level paraphyly and polyphyly: frequency, causes, and
consequences, with insights from animal mitochondrial DNA. Annu. Rev. Ecol. Evol.
Syst. 34:397-423.
89
Gelman A., Meng X. 1998. Simulating normalizing constants: from importance sampling to
bridge sampling to path sampling. Statist. Sci. 13:163-185.
Geyer C.J. 1991. Markov chain Monte Carlo maximum likelihood. In: Keramides E.M., editor.
Computing Science and Statistics: Proceedings of the 23rd Symposium on the
Interface. Fairfax Station, USA, Interface Foundation. p. 156-163.
Gilks W.R., Roberts G.O. 1996. Strategies for improving MCMC. In: Gilks W.R., Richardson S.,
Spiegelhalter, editors. Markov chain Monte Carlo in Practice. London, UK, Chapman
and Hall. p. 89-114.
Giroux M., Pape T., Wheeler T. 2010. Towards a phylogeny of the flesh flies (Diptera:
Sarcophagidae): Morphology and phylogenetic implications of the acrophallus in the
subfamily Sarcophaginae. Zool. J. Linnean Soc. 158:740-778.
Greenberg B. 1991. Flies as forensic indicators. J. Med. Entomol. 28:565-577.
Grummer J.A., Bryson R.W., Jr., Reeder T.W. 2014. Species delimitation using Bayes factors:
simulation and application to the Sceloporus scalaris species group (Squamata:
Phrynosomatidae). Syst. Biol. 63:119-133.
Guo Y.D., Cai J.F., Li X., Xiong F., Su R.N., Chen F.L., Liu Q.L., Wang X.H., Chang Y.F., Zhong M.,
Wang X., Wen J.F. 2010. Identification of the forensically important sarcophagid flies
Boerttcherisca peregrina, Parasarcophaga albiceps and Parasarcophaga dux
(Diptera: Sarcophagidae) based on COII gene in China. Trop. Biomed. 27:451-460.
Hamilton C.A., Formanowicz D.R., Bond J.E. 2011. Species delimitation and phylogeography
of Aphonopelma hentzi (Araneae, Mygalomorphae, Theraphosidae): cryptic diversity
in North American tarantulas. PLoS ONE. 6:e26207.
90
Harrington R.C., Near T.J. 2012. Phylogenetic and Coalescent Strategies of Species
Delimitation in Snubnose Darters (Percidae: Etheostoma). Syst. Biol. 61:63-79.
Hasegawa M., Kishino H., Yano T. 1985. Dating of the human-ape splitting by a molecular
clock of mitochondrial DNA. J. Mol. Evol. 22:160-174.
Heled J., Drummond A.J. 2010. Bayesian inference of species trees from multilocus data.
Mol. Biol. Evol. 27:570-580.
Hey J. 2010. Isolation with migration models for more than two populations. Mol. Biol. Evol.
27:905-920.
Hey J. 2011. Documentation for IMa2. <http://genfaculty.rutgers.edu/hey/software>.
Hudson R.R. 1991. Gene genealogies and the coalescent process. Oxf. Surv. Evol. Biol. 7:1-
44.
ICZN (International Commission on Zoological Nomenclature) (1999) International code of
zoological nomenclature, 4th edn. London, UK: International Trust for Zoological
Nomenclature.
Jakobsson M., Rosenberg N.A. 2007. CLUMPP: a cluster matching and permutation program
for dealing with label switching and multimodality in analysis of population
structure. Bioinformatics. 23:1801-1806.
Jordaens K., Sonet G., Richet R., Dupont E., Braet Y., Desmyter S. 2013. Identification of
forensically important Sarcophaga species (Diptera: Sarcophagidae) using the
mitochondrial COI gene. Int. J. Legal. Med. 127:491-504.
Kass R.E., Raftery A.E. 1995. Bayes Factors. J. Am. Statist. Assoc. 90:773-795.
91
Keck B.P., Near T.J. 2010. Geographic and temporal aspects of mitochondrial replacement in
Nothonotus darters (Teleostei: Percidae: Etheostomatinae). Evolution. 64:1410-
1428.
Keith R., Hedin M. 2012. Extreme mitochondrial population subdivision in southern
Appalachian paleoendemic spiders (Araneae: Hypochilidae: Hypochilus), with
implications for species delimitation. J. Arachnology. 40:167-181.
Klingman J.F.C. 1982. The Coalescent. Stoch. Proc. Appl. 13:235-248.
Knowles L.L., Carstens B.C. 2007. Delimiting species without monophyletic gene trees. Syst.
Biol. 56:887-895.
Kutty S.N., Pape T., Wiegmann B.M., Meier R. 2010. Molecular phylogeny of the Calypratae
(Diptera: Cyclorrhapha) with an emphasis on the superfamily Oestroidea and the
position of the Mystacinobiidae and McAlpine’s fly. Syst. Entomol. 35:614-635.
Lanfear R., Calcott B., Ho S.Y.W., Guindon S. 2012. PartitionFinder: Combined selection of
partitioning schemes and substitution models for phylogenetic analyses. Mol. Biol.
Evol. 29:1695-1701.
Lartillot N., Philippe H. 2006. Computing Bayes factors using thermodynamic integration.
Syst. Biol. 55:195-207.
Leaché A.D., Fujita M.K. 2010. Bayesian species delimitation in West African forest geckos
(Hemidactylus fasciatus). Proc. Royal Soc. B. 277:3071-3077.
Leaché A.D., Harris R.B., Rannala B., Yang Z. 2014. The influence of gene flow on species tree
estimation: a simulation study. Syst. Biol. 63:17-30.
Librado P., Rozas J. 2009. DnaSP v5: A software for comprehensive analysis of DNA
polymorphism data. Bioinformatics 25:1451-1452.
92
Llopart A., Lachaise D., Coyne J.A. 2005. Multilocus analysis of introgression between
sympatric sister species of Drosophila: Drosophila yakuba and D. santomea. Genetics.
171:197-210.
Lohse K. 2009. Can mtDNA barcodes be used to delimit species? A response to Pons et al.
(2006). Syst. Biol. 58:439-442.
Lopes H.S. 1982. The importance of the mandible and clypeal arch of the first instar larvae
in the classification of the Sarcophagidae (Diptera). Rev. Bras. Biol. 26:293-326.
Lowenstein J.H., Amato G., Kolokotronis S.-O. 2009. The Real maccoyii: Identifying Tuna
Sushi with DNA Barcodes – Contrasting Characteristic Attributes and Genetic
Distances. PLoS ONE. 4: e7866.
Maddison W.P., Maddison D.R. 2011. ‘Mesquite 2.73: a modular system for evolutionary
analysis.’ Available from [http://mesquiteproject.org].
Mayr E. 1942. Systematics and the Origin of Species. Columbia University Press, New York.
Mayr E. 1963. Animal Species and Evolution. Harvard University Press, Cambridge,
Massachusetts.
Mayer D.G., Atzeni M.G. 1993. Estimation of Dispersal Distances for Cochliomyia
hominivorax (Diptera: Calliphoridae). Environ. Entomol. 22:368-374.
McGuire G., Wright G., Prentice M.J. 1997. A graphical method for detecting recombination
in phylogenetic data sets. Mol. Biol. Evol. 14:1125-1131.
Meiklejohn K.A., Wallman J.F., Dowton M. 2011. DNA-based identification of forensically
important Australian Sarcophagidae (Diptera). Int. J. Leg. Med. 125:27-32.
93
Meiklejohn K.A., Wallman J.F., Cameron S.L., Dowton M. 2012. Comprehensive evaluation of
DNA barcoding for the molecular species identification of forensically important
Australian Sarcophagidae (Diptera). Invert. Syst. 26:515-525.
Meiklejohn K.A., Wallman J.F., Dowton M. 2013a. DNA Barcoding Identifies all Immature
Life Stages of a Forensically Important Flesh Fly (Diptera: Sarcophagidae). J.
Forensic. Sci. 58:184-187.
Meiklejohn K.A., Wallman J.F., Pape T., Cameron S.L., Dowton M. (2013b) Utility of COI, CAD
and morphological data for resolving relationships within the genus Sarcophaga
(sensu lato) (Diptera: Sarcophagidae): A preliminary study. Mol. Phylo. Evol.
69:133-141.
Milne I., Lindner D., Bayer M., Husmeier D., McGuire G., Marshall D.F., Wright F. 2009.
TOPALi v2: a rich graphical interface for evolutionary analyses of multiple
alignments on HPC clusters and multi-core desktops. Bioinformatics. 25:126-127.
Monaghan M.T., Wild R., Elliot M., Fujisawa T., Balke M., Inward D.J.G., Lees D.C.,
Ranaivosolo R., Eggleton P., Barraclough T.G., Vogler A.P. 2009. Accelerated Species
Inventory on Madagascar Using Coalescent-Based Models of Species Delineation.
Syst. Biol. 58:298-311.
Nee S., May R.M., Harvey P.H. 1994. The reconstructed evolutionary process. Phil. Trans. R.
Soc. B. 344:305-11.
Nielsen R., Wakeley J. 2001. Distinguishing migration from isolation: a Markov chain Monte
Carlo approach. Genetics. 158:885-896.
94
Niemiller M.L., Fitzpatrick B.M., Miller B.T. 2008. Recent divergence-with-gene-flow in
Tennessee cave salamanders (Plethonodontidae: Gyrinophilus) inferred from gene
genealogies. Mol. Ecol. 17:2258-2275.
Niemiller M.L., Near T.J., Fitzpatrick B.M. 2012. Delimiting species using multilocus data:
diagnosing cryptic diversity in the southern cavefish, Typhlichthys subterraneus
(Teleostei: Amblyopsidae). Evolution. 66:846-866.
Pape T. 1994. The world Blaesoxipha Loew, 1861 (Diptera: Sarcophagidae). Entomol.
Scand. Suppl. 45:1-247.
Pape T. 1996. Catalogue of the Sarcophagidae of the World (Insecta: Diptera), Memoirs on
Entomology International, Vol. 8. Associated Publishers, Gainesville, FL.
Pape T., Dahlem G.A. 2010. Sarcophagidae (Flesh Flies). In: Brown B.V., et al., editors.
Manual of Central American Diptera, Volume 2. Ottawa, Ontario, CAN., NRC Research
Press p. 1313-1335.
Pons J., Barraclough T.G., Gomez-Zurita J., Cardoso A., Duran D.P., Hazell S., Kamoun S.,
Sumlin W.D., Vogler A.P. 2006. Sequence-based species delimitation for the DNA
taxonomy of undescribed insects. Syst. Biol. 55:595-609.
Pons J., Fujisawa T., Claridge E.M., Savill R.A., Barraclough T.G., Vogler A.P. 2011. Deep
mtDNA subdivision within Linnean species in an endemic radiation of tiger beetles
from New Zealand (genus Neocicindela). Mol. Phyl. Evol. 59:251-262.
Pritchard J.K., Stephens M., Donnelly P. 2000. Inference of population structure using
multilocus genotype data. Genetics. 155:945-959.
95
Pritchard J.K., Wen X., Falush D. 2010. Documentation for structure software: Version 2.3,
Available from [http://pritchardlab.stanford.edu/structure_software/
release_versions/v2.3.4/structure_doc.pdf] Last accessed 01/27/2015.
Rambaut A., Suchard M.A., Xie D., Drummond A.J. 2014. Tracer v1.6, Available
from [http://beast.bio.ed.ac.uk/Tracer].
Reid N.M., Carstens B.C. 2012. Phylogenetic estimation error can decrease the accuracy of
species delimitation: a Bayesian implementation of the general mixed Yule-
coalescent model. BMC Evol. Biol. 12:196.
Robineau-Desvoidy J.B. 1863. Histoire naturelle des Dipteres des environs de Paris. Vol. 2,
pp. 920. (Paris, France).
Ronquist F., Huelsenbeck J.P. 2003. MrBayes 3: Bayesian phylogenetic inference under
mixed models. Bioinformatics. 19:1572-1574.
Rosenberg N.A., Nordborg M. 2002. Genealogical trees, coalescent theory and the analysis
of genetic polymorphisms. Nat. Rev. Genet. 3:380-390.
Rosenberg N.A. 2004. Distruct: a program for the graphical display of population structure.
Mol. Ecol. Notes. 4:137-138.
Satler J.D., Carstens B.C., Hedin M. 2013. Multilocus Species Delimitation in a Complex of
Morphologically Conserved Trapdoor Spiders (Mygalomorphae, Antrodiaetidae,
Aliatypus). Syst. Biol. 62:805-823.
Shaw K.L. 2002. Conflict between nuclear and mitochondrial DNA phylogenies of a recent
species radiation: what mtDNA reveals and conceals about modes of speciation in
Hawaiian crickets. Proc. Royal Soc. B. 99:16122-16127.
96
Song H., Buhay J.E., Whiting M.F., Crandall K.A. 2008. Many species in one: DNA barcoding
overestimates the number of species when nuclear mitochondrial pseudogenes are
coamplified. Proc. Natl. Acad. Sci. USA. 105:13486-13491.
Spillings B.L., Brooke B.D., Koekemoer L.L., Chiphwanya J., Coetzee M., Hunt R.H. 2009. A
new species concealed by Anopheles funestus Giles, a major malaria vector in Africa.
Am. J. Trop. Med. Hyg. 81:510-515.
Stamatakis A. 2014. RAxML Version 8: A tool for Phylogenetic Analysis of Large
Phylogenies. Bioinformatics. 30:1312-1313.
Stamper T., Dahlem G.A., Cookman C., DeBry R.W. 2013. Phylogenetic relationships of flesh
flies in the subfamily Sarcophaginae based on three mtDNA fragments (Diptera:
Sarcophagidae). Syst. Entomol. 38:35-44.
Stephens M., Donnelly P. 2003. Ancestral inference in population genetics models with
selection. Aust. NZ. J. Stat. 45:901-931.
Stephens M., Smith N.J., Donnelly P. 2001. A new statistical method for haplotype
reconstruction from population data. Am. J. Hum. Genet. 68:978-989.
Sukumaran J., Holder M.T. 2010. DendroPy: A Python library for phylogenetic computing.
Bioinformatics. 26:1569-1571.
Tan S.H., Rizman-Idid M., Mohd-Aris E., Kurahashi H., Mohamed Z. 2010. DNA-based
characterization and classification of forensically important flesh flies (Diptera:
Sarcophagidae) in Malaysia. Forensic Sci. Int. 199:43-49.
Wakely J. 2006. Coalescent theory: an introduction. Roberts and Co., Greenwood Village,
Colorado.
97
Wiens J.J. 2007. Species Delimitation: New Approaches for Discovering Diversity. Syst. Biol.
56:875-878.
Wright S. 1931. Evolution in Mendelian populations. Genetics. 16:97-159.
Wong E.S., Dahlem G.A., Stamper T.I., DeBry R.W. 2015. Discordance between
morphological species identification and mtDNA phylogeny in the flesh fly genus
Ravinia (Diptera: Sarcophagidae). Invert. Syst. 29:1-11.
Xie W., Lewis P.O., Fan Y., Kuo L., Chen M.-H. 2011. Improving marginal likelihood
estimation for Bayesian phylogenetic model selection. Syst. Biol. 60:150-160.
Yang Z., Rannala B. 2010. Bayesian species delimitation using multilocus sequence data.
Proc. Natl. Acad. Sci. USA. 107:9264-9269.
Zhang C., Zhang D.X., Zhu T., Yang Z. 2011. Evaluation of a Bayesian coalescent method of
species delimitation. Syst. Biol. 60:747-761.
98
3.10 SUPPLEMENTAL MATERIALS
APPENDIX: S3.1. Primer Information and Amplification Procedure:
Mitochondrial gene fragments COI and COII were amplified using primers and
specifications given in DeBry et al. (2013) and Stamper et al. (2013), respectively.
Nuclear gene fragments for elongation factor 1 alpha (EF-1a), white (Wt), period
(Per), glucose-6-phosphate dehydrogenase (G6PD), phosphoenolpyruvate carboxykinase
(PEPCK). All fragments except those listed were generated directly from published fly
sequences*. We have listed GenBank accession numbers and citations of the flies used to
generate these new primers.
Table S-3.1. Primer sequence information
Gene Name Sequence Citation EF-1a
EF1a-55f 5'-GGT-ATC-ACC-ATT-GAT-ATT-GCT-TTG-TGG-3' Stamper 2008
EF1aEWF 5'-ATC-ATT_GAT-GCT-CCT-GGT-CA-3' This Study
EF1aEWR 5'-TTG-AAA-CCA-ACG-TTG-TCA-CC-3' This Study
EF1a836R 5'-CAG-CAG-CAC-CTT-TAG-GTG-GGC-TAG-CCT-T-3' Stamper 2008
White
WECF21 5'-GTT-TGT-GGC-GTA-GCC-TAT-CC-3' Ready et al. 2009
WECF195 5'-CGC-GCA-CTA-TGA-CTC-AAA-AA-3' This Study
WECR459 5'-TTG-CCA-CGT-TGG-GAT-AAT-TT-3' This Study
WECR12 5'-AAT-GTC-ACT-CTA-CCY-TCG-GC-3' Ready et al. 2009
Period Per1607F 5'-GCG-AGA-TCT-CCC-CGC-AYC-AYG-AYT-A-3' This Study
PerEWF 5'-TAT-AGT-GGC-CCA-GGG-CAT-3' This Study
2007F 5'-TCG-CGA-ATC-CCG-TGG-ACG-AA-3' This Study
PerEWR 5'-CCA-GCC-ATA-TAT-TGA-AGA-3' This Study
2574R 5'-CAT-GGC-AGC-GGC-AGG-ATG-AGT-3' This Study
G6PD
G6PD40F 5'-TCC-ACA-TGG-AAT-CGT-GAA-AA-3' This Study
G6PD87F 5'-GGA-ACC-TTT-CGG-TAC-TGA-GG-3' This Study
G6PD599R 5'-GGC-GAT-TTG-GTC-ATC-ATT-TT-3' This Study
G6PD690R 5'-ACG-TTC-ATA-GGC-ATC-GGG-TA-3' This Study
PEPCK
PEPCK37F 5'-AGT-TTG-GCC-GAA-GGT-GTT-C-3' This Study
PEPCK427R 5'-GTA-CTT-GGC-CAC-GAG-ATT-CC-3' This Study
PEPCK455R 5'-ACG-AAC-CAG-TTG-ACG-TGG-A-3' This Study
99
Fly sequences used for generating primers*: Elongation factor 1 alpha: -Amplified sequences were obtained from those specimens that worked with Stamper
(2008) primer set EF1a55f – EF1a836R. Subsequent primers were then generated from sequenced Ravinia individuals.
White: -Amplified sequences were obtained from those specimens that worked with Ready et al.
(2009) primer set WECF21-WECR12. Subsequent primers were then generated from sequenced Ravinia individuals.
Period: -Cochliomyia macellaria (GenBank Accession Number: KC178068; Wiegmann et al. 2011). -Musca domestica (GenBank Accession Number: KC178071; Wiegmann et al. 2011). -Sarcophaga bullata (GenBank Accession Number: KC178072; Wiegmann et al. 2011). Glucose-6-phosphate dehydrogenase: -Cochliomyia macellaria (GenBank Accession Number: KC178100; Wiegmann et al. 2011). -Musca domestica (GenBank Accession Number: KC178102; Wiegmann et al. 2011). -Sarcophaga bullata (GenBank Accession Number: KC178103; Wiegmann et al. 2011). Phosphoenolpyruvate carboxykinase: -Cochliomyia macellaria (GenBank Accession Number: KC177343; Wiegmann et al. 2011). -Musca domestica (GenBank Accession Number: KC177346; Wiegmann et al. 2011). -Sarcophaga bullata (GenBank Accession Number: KC177347; Wiegmann et al. 2011). Amplification Protocols:
All nuclear gene fragments were first amplified using standard PCR, if this failed then nested PCR was used. All primers for both the first and the second rounds of PCR were able to be amplified using the following parameters: 95˚C for 3 min, followed by 35 cycles of: 94˚C for 1 s, 60˚C for 1 s, 68˚C for 20 s. PCR schemes were as follows: EF-1a:
1st Attempt PCR: EF1aEWF-EF1a836R @ Annealing 60˚C 2nd Attempt PCR: EF1a55f-EF1a836R @ Annealing 60˚C 3rd Attempt Nested PCR: [EF1a55f-EF1a836R] → EF1aEWF-EF1a836R @ Annealing 60˚C 4th Attempt Nested PCR: [EF1a55f-EF1a836R] → EF1a55f-EF1aEWR @ Annealing 60˚C
White:
1st Attempt PCR: WECF21-WECR12 @ Annealing 60˚C 2nd Attempt Nested PCR: [WECF21-WECR12] → WECF21-WECR459 @ Annealing 60˚C 3rd Attempt Nested PCR: [WECF21-WECR12] → WECF195-WECR12 @ Annealing 60˚C
100
Period:
1st Attempt Nested PCR: [1607-2574] → PerEWF-2574 @ Annealing 60˚C 2nd Attempt Nested PCR: [1607-2574] → 2007-2574 @ Annealing 60˚C 3rd Attempt Nested PCR: [1607-2574] → PerEWF-PerEWR @ Annealing 60˚C
G6PD:
1st Attempt Nested PCR: [G6PD40F-G6PD690R] → G6PD87F-G6PD599R @ Annealing 60˚C
PEPCK:
1st Attempt PCR: PEPCK37F-PEPCK455R @ Annealing 60˚C 2nd Attempt Nested PCR: [PEPCK37F-PEPCK455R] → PEPCK37F-PEPCK427R @
Annealing 60˚C
DeBry R.W., Timm A., Wong E.S., Stamper T., Cookman C., Dahlem G.A. 2013. DNA-Based Identification of Forensically Important Lucilia (Diptera: Calliphoridae) in the Continental United States*. J. Forensic Sci. 58(1):73-78.
Ready P.D., Testa J.M., Wardhana A.H., Al-Izzi M., Khalaj M., Hall M.J.R. 2009. Phylogeography and recent emergence of the Old World screwworm fly, Chrysomya bezziana, based on mitochondrial and nuclear gene sequences. Med. Vet. Entomol. 23:43-50.
Stamper T. 2008. Improving the Accuracy of Postmortem Interval Estimations Using Carrion Flies (Diptera: Sarcophagidae, Calliphoridae and Muscidae). Ph.D. Dissertation. University of Cincinnati, Cincinnati, Ohio, USA.
Stamper T., Dahlem G.A., Cookman C., DeBry R.W. 2013. Phylogenetic relationships of flesh flies in the subfamily Sarcophaginae based on three mtDNA fragments (Diptera: Sarcophagidae). Syst. Entomol.. 38(1):35-44.
Wiegmann B.M., Trautwein M.D., Winkler I.S., Barr N.B., Kim J.-W., Lambkin C., Bertone M.A., Cassel B.K., Bayless K.M., Heimberg A.M., Wheeler B.M., Peterson K.J., Pape T., Sinclair B.J., Skevington J.H., Blagoderov V., Caravas J., Kutty S.N., Schmidt-Ott U., Kampmeier G.E., Thompson C.F., Grimaldi D.A., Beckenbach A.T., Courtney G.W., Friedrich M., Meier R., Yeates D.K. 2011. Episodic radiations in the fly tree of life. Proc.Natl. Acad. Sci.USA. 108(14):5690-5695.
101
APPENDIX: S3.2. Best Model Selection and Partition Schemes:
Bayesian Inference Analyses (MrBayes)
R. anxia and R. querula data set COI: COI_pos1: GTR; COI_pos2: HKY; COI_pos3: HKY, Scheme lnL = -2496.96783; Scheme AIC = 5271.93566
R. anxia and R. querula data set COII: COII_pos1: HKY + I; COII_pos2: HKY; COII_pos3: HKY, Scheme lnL = -1155.25889; Scheme AIC = 2582.51778
R. anxia and R. querula data set EF1a: EF1a_pos1: HKY; EF1a_pos2: F81 + I; EF1a_pos3: HKY, Scheme lnL = -1146.63463; Scheme AIC = 2563.26926
R. anxia and R. querula data set White: White_pos1: K80 + I; White_pos2: HKY + I + G; White_pos3: HKY + I + G, Scheme lnL = -1293.58635; Scheme AIC = 2861.1727
R. anxia and R. querula data set Period: (Per_pos1, Per_pos2): GTR + G; Per_pos3: HKY, Scheme lnL = -935.49325; Scheme AIC = 2140.9865
R. anxia and R. querula data set PEPCK: (PEPCK_pos1, PEPCK_pos2): K80 + I; PEPCK_pos3: HKY + G, Scheme lnL = -724.60091, Scheme AIC = 1707.20182
R. anxia and R. querula data set G6PD: G6PD_pos1: HKY; G6PD_pos2: F81; G6PD_pos3: HKY + G, Scheme lnL = -1273.5593; Scheme AIC = 2817.1186
R. anxia and R. querula All Genes data set: PartitionFinder Input: All genes, all codon position. Greedy algorithm run to
determine best-fit models and partitions; resulted in 19 subsets. Scheme lnL = -10311.25024; Scheme AIC = 2122.50048
Subset 1: COI_pos1 = GTR + I Subset 2: COI_pos2 = HKY + I Subset 3: (COI_pos3, COII_pos3) = HKY Subset 4: COII_pos1 = HKY + I Subset 5: COII_pos2 = HKY + I Subset 6: EF1a_pos1 = HKY + I Subset 7: EF1a_pos2 = GTR + I + G Subset 8: EF1a_pos3 = HKY + I + G Subset 9: White_pos1 = SYM + I Subset 10: White_pos2 = HKY + I + G Subset 11: White_pos3 = GTR + I + G Subset 12: (Per_pos1, Per_pos2) = GTR + G Subset 13: Per_pos3 = HKY + G Subset 14: G6PD_pos1 = HKY Subset 15: G6PD_pos2 = F81 Subset 16: G6PD_pos3 = GTR + G Subset 17: PEPCK_pos1 = HKY + I + G Subset 18: PEPCK_pos2 = JC
Subset 19: PEPCK_pos3 = GTR + I + G
102
Maximum Likelihood Analyses (RAxML)
R. anxia and R. querula COI: 1st Codon Pos., 2nd Codon Pos., 3rd Codon Pos. = GTR + G. Scheme lnL = -2491.96537; Scheme AIC = 5283.93074
R. anxia and R. querula COII: (1st Codon Pos., 2nd Codon Pos.), 3rd Codon Pos. = GTR + G. Scheme lnL = -1159.64217; Scheme AIC = 2467.28434
R. anxia and R. querula EF1a: 1st Codon Pos., (2nd Codon Pos., 3rd Codon Pos.) = GTR + G. Scheme lnL = -1151.39379; Scheme AIC 2570.78758
R. anxia and R. querula G6PD: 1st Codon Pos., 2nd Codon Pos., 3rd Codon Pos. = GTR + G. Scheme lnL = -1258.15657; Scheme AIC = 2812.31314
R. anxia and R. querula PEPCK: (1st Codon Pos., 2nd Codon Pos.), 3rd Codon Pos. = GTR + G. Scheme lnL = -717.65079; Scheme AIC = 1711.30158
R. anxia and R. querula Period: (1st Codon Pos., 2nd Codon Pos.), 3rd Codon Pos. = GTR + G. Scheme lnL =-923.40787; Scheme AIC = 2114.81574
R. anxia and R. querula White: (1st Codon Pos., 2nd Codon Pos.), 3rd Codon Pos. = GTR + I + G. Scheme lnL = -1177.41104; Scheme 2634.82208
R. anxia and R. querula All Genes: PartitionFinder Input: All genes, all codon position.
Greedy algorithm run to determine best-fit model and partitions; resulted in 15 subsets. Scheme lnL = -10303.77703; Scheme AIC = 21173.55406
Subset 1: (COII_pos1, COI_pos1) = GTR + I + G Subset 2: (COII_pos2, COI_pos2) = GTR + I + G Subset 3: (COII_pos3, COI_pos3) = GTR + I + G Subset 4: (EF1a_pos1, White_pos2) = GTR + I + G Subset 5: EF1a_pos2 = GTR + I + G Subset 6: EF1a_pos3 = GTR + I + G Subset 7: (PEPCK_pos2, White_pos1) = GTR + I + G Subset 8: White_pos3 = GTR + I + G Subset 9: (Per_pos1, Per_pos2) = GTR + I + G Subset 10: Per_pos3 = GTR + I + G Subset 11: G6PD_pos1 = GTR + I + G Subset 12: G6PD_pos2 = GTR + I + G Subset 13: G6PD_pos3 = GTR + I + G Subset 14: PEPCK_pos1 = GTR + I + G Subset 15: PEPCK_pos3 = GTR + I + G
*BEAST Analysis
COICOII: (COI_pos1, COI_pos2, COI_pos3, COII_pos1, COII_pos2, COII_pos3) = GTR + I + G, Scheme lnL = -7857.12668; Scheme AIC = 16184.25336
EF1a: (EF1a_pos1, EF1a_pos2, EF1a_pos3) = SYM + I, Scheme lnL = -1436.67395; Scheme AIC = 3335.3479
G6PD: (G6PD_pos1, G6PD_pos2, G6PD_pos3) = TN93 (TrN + I + G), Scheme lnL = -2167.08558; Scheme AIC = 4798.17116
PEPCK: (PEPCK_pos1, PEPCK_pos2, PEPCK_pos3) = SYM + G, Scheme lnL = -1289.59556; Scheme AIC = 3041.19112
103
Period: (Period_pos1, Period_pos2, Period_pos3) = HKY + G, Scheme lnL = -1902.40222; Scheme AIC = 4264.80444
White: (White_pos1, White_pos2, White_pos3) = GTR + I + G, Scheme lnL = -2327.89807; Scheme AIC = 5125.79614
104
SUPP
LEM
ENTA
L TA
BLES
Tabl
e S3
.2. L
ist o
f spe
cim
ens,
spec
imen
iden
tific
atio
n (I
.D.)
code
s, co
llect
ion
data
, and
Gen
Bank
Acc
essi
on n
umbe
rs.
Spec
ies
I.D.
Loca
lity
Sex
COI
COII
EF
-1a
Whi
te
Peri
od
G6PD
PE
PCK
Ravi
nia
anxi
a (W
alke
r)
AE79
Si
skiy
ou, C
A M
KR
6083
07
KR67
7021
KR
6083
24
KR67
6632
KR
6768
14
KR67
6761
KR
6767
08
Ravi
nia
anxi
a (W
alke
r)
AE80
Si
skiy
ou, C
A M
KR
6083
08
KR67
7022
KR
6083
25
KR67
6633
KR
6768
15
KR67
6762
KR
6767
09
Ravi
nia
anxi
a (W
alke
r)
AE81
Si
skiy
ou, C
A M
JQ
8070
69
KJ58
6403
KR
6083
26
KR67
6634
KR
6768
16
KR67
6763
KR
6767
10
Ravi
nia
anxi
a (W
alke
r)
AH52
Je
ffers
on, O
R F
KJ58
6344
KJ
5864
05
KR60
8328
KR
6766
36
KR67
6817
KR
6767
64
KR67
6711
Ravi
nia
anxi
a (W
alke
r)
AJ18
Je
ffers
on, O
R M
KJ
5863
45
KJ58
6406
KR
6083
29
KR67
6637
KR
6768
18
KR67
6765
KR
6767
12
Ravi
nia
anxi
a (W
alke
r)
AK25
Je
ffers
on, O
R F
KJ58
6346
KJ
5864
07
KR60
8330
KR
6766
38
KR67
6819
KR
6767
66
KR67
6713
Ravi
nia
anxi
a (W
alke
r)
AW48
Gr
ant,
NM
F
KJ58
6351
KJ
5864
14
KR60
8331
KR
6766
39
KR67
6820
KR
6767
67
KR67
6714
Ravi
nia
anxi
a (W
alke
r)
AW70
Gr
ant,
NM
M
KR
6083
09
KR67
7023
KR
6083
32
KR67
6640
KR
6768
21
KR67
6768
KR
6767
15
Ravi
nia
flori
dens
is (A
ldri
ch)
AA01
Du
val,
FL
M
JQ80
7076
KJ
5863
49
KR60
8346
KR
6766
50
KR67
6824
KR
6767
72
KR67
6719
Ravi
nia
quer
ula
(Wal
ker)
AF
03
Sisk
iyou
, CA
M
KJ58
6379
KJ
5864
66
KR60
8372
KR
6766
76
KR67
6838
KR
6767
88
KR67
6735
Ravi
nia
quer
ula
(Wal
ker)
AF
16
Clac
kam
as, O
R M
KJ
5863
80
KJ58
6467
KR
6083
73
KR67
6677
KR
6768
39
KR67
6789
KR
6767
36
Ravi
nia
quer
ula
(Wal
ker)
AG
28
Jeffe
rson
, OR
M
KJ58
6381
KJ
5864
68
KR60
8374
KR
6766
78
KR67
6840
KR
6767
90
KR67
6737
Ravi
nia
quer
ula
(Wal
ker)
AG
30
Jeffe
rson
, OR
M
KJ58
6382
KJ
5864
69
KR60
8375
KR
6766
79
KR67
6841
KR
6767
91
KR67
6738
Ravi
nia
quer
ula
(Wal
ker)
AG
31
Jeffe
rson
, OR
M
KJ58
6383
KJ
5864
70
KR60
8376
KR
6766
80
KR67
6842
KR
6767
92
KR67
6739
Ravi
nia
quer
ula
(Wal
ker)
AG
32
Jeffe
rson
, OR
M
KJ58
6384
KJ
5864
71
KR60
8377
KR
6766
81
KR67
6843
KR
6767
93
KR67
6740
Ravi
nia
quer
ula
(Wal
ker)
AG
33
Jeffe
rson
, OR
M
KJ58
6385
KJ
5864
72
KR60
8378
KR
6766
82
KR67
6844
KR
6767
94
KR67
6741
Ravi
nia
quer
ula
(Wal
ker)
AH
42
Croo
k, O
R M
KJ
5863
86
KJ58
6473
KR
6083
79
KR67
6683
KR
6768
45
KR67
6795
KR
6767
42
Ravi
nia
quer
ula
(Wal
ker)
AH
51
Croo
k, O
R F
KJ58
6387
KJ
5864
74
KR60
8380
KR
6766
84
KR67
6846
KR
6767
96
KR67
6743
Ravi
nia
quer
ula
(Wal
ker)
AJ
19
Jeffe
rson
, OR
M
KJ58
6388
KJ
5864
75
KR60
8381
KR
6766
85
KR67
6847
KR
6767
97
KR67
6744
Ravi
nia
quer
ula
(Wal
ker)
AJ
23
Jeffe
rson
, OR
F KJ
5863
89
KJ58
6476
KR
6083
82
KR67
6686
KR
6768
48
KR67
6798
KR
6767
45
Ravi
nia
quer
ula
(Wal
ker)
AK
04
Jeffe
rson
, OR
M
KR60
8316
KR
6770
28
KR60
8383
KR
6766
87
KR67
6489
KR
6767
99
KR67
6746
Ravi
nia
quer
ula
(Wal
ker)
AK
05
Jeffe
rson
, OR
M
KJ58
6390
KJ
5864
77
KR60
8384
KR
6766
88
KR67
6850
KR
6768
00
KR67
6747
Ravi
nia
quer
ula
(Wal
ker)
AL
32
Rabu
n, G
A M
KJ
5863
91
KJ58
6478
KR
6083
85
KR67
6689
KR
6768
51
KR67
6801
KR
6767
48
Ravi
nia
quer
ula
(Wal
ker)
AS
25
Mar
atho
n, W
I F
KJ58
6392
KJ
5864
79
KR60
8386
KR
6766
90
KR67
6852
KR
6768
02
KR67
6749
105
Ravi
nia
quer
ula
(Wal
ker)
AW
30
Gran
t, N
M
M
KR60
8317
KR
6770
29
KR60
8387
KR
6766
91
KR67
6853
KR
6768
03
KR67
6750
Ravi
nia
quer
ula
(Wal
ker)
AZ
61
Blan
d, V
A M
JQ
8071
01
KJ58
6480
KR
6083
88
KR67
6692
KR
6768
54
KR67
6804
KR
6767
51
Ravi
nia
quer
ula
(Wal
ker)
BA
16
Gran
t, N
M
M
JQ80
7102
KJ
5864
81
KR60
8389
KR
6766
93
KR67
6855
KR
6768
05
KR67
6752
Ravi
nia
quer
ula
(Wal
ker)
BI
26
Utah
, UT
M
KR60
8318
---
------
---
KR60
8390
KR
6766
94
KR67
6856
KR
6768
06
KR67
6753
Ravi
nia
quer
ula
(Wal
ker)
BI
29
Utah
, UT
M
KR60
8319
KR
6770
30
------
------
- KR
6766
95
------
------
- KR
6768
07
KR67
6754
Ravi
nia
quer
ula
(Wal
ker)
BI
46
Utah
, UT
M
KR60
8320
KR
6770
31
KR60
8391
KR
6766
96
KR67
6857
KR
6768
08
KR67
6755
Ravi
nia
quer
ula
(Wal
ker)
E5
6 W
ashi
ngto
n, T
N
M
KR60
8321
KR
6770
32
KR60
8392
KR
6766
97
KR67
6858
KR
6768
09
KR67
6756
106
Table S3.3. BP&P results with guide tree parameters with adjusted population size (ϴ) and
root ages (τo) using a gamma distribution (α, β).
Note: *All nodes supported by a posterior probability of 1.00
Population Size (ϴ) Root age (τo) I.D. *Model Probability Final lnL α β α β
Morphology Guide Tree Large ancestral populations and deep divergences
1 10 1 10 BP1 P[3]=1.00 -10558.3
Small ancestral populations and shallow divergences
2 1000 2 1000 BP2 P[3]=1.00 -10565.0
Large ancestral populations with shallow divergences
1 10 2 1000 BP3 P[3]=1.00 -10557.3
Small ancestral populations with large divergence
2 1000 1 10 BP4 P[3]=1.00 -10563.9
Mitochondrial Guide Tree Large ancestral populations and deep divergences
1 10 1 10 BP1 P[4]=1.00 -10557.7
Small ancestral populations and shallow divergences
2 1000 2 1000 BP2 P[4]=1.00 -10559.1
Large ancestral populations with shallow divergences
1 10 2 1000 BP3 P[4]=1.00 -10556.9
Small ancestral populations with large divergence
2 1000 1 10 BP4 P[4]=1.00 -10560.3
DNA + Morphology Guide Tree Large ancestral populations and deep divergences
1 10 1 10 BP1 P[5]=1.00 -10559.1
Small ancestral populations and shallow divergences
2 1000 2 1000 BP2 P[5]=1.00 -10560.7
Large ancestral populations with shallow divergences
1 10 2 1000 BP3 P[5]=1.00 -10558.5
Small ancestral populations with large divergence
2 1000 1 10 BP4 P[5]=1.00 -10561.1
107
SUPPLEMENTAL FIGURES
Figure S3.1. Inferred COI gene-tree using maximum likelihood and Bayesian inference.
Support values for nodes are ML-bootstrap support (%)/BI-posterior probabilities.
109
Figure S3.2. Inferred COII gene-tree using maximum likelihood and Bayesian inference.
Support values for nodes are ML-bootstrap support (%)/BI-posterior probabilities.
111
Figure S3.3. Inferred EF-1a gene-tree using maximum likelihood and Bayesian inference.
Support values for nodes are ML-bootstrap support (%)/BI-posterior probabilities.
113
Figure S3.4. Inferred White gene-tree using maximum likelihood and Bayesian inference.
Support values for nodes are ML-bootstrap support (%)/BI-posterior probabilities.
115
Figure S3.5. Inferred Period gene-tree using maximum likelihood and Bayesian inference.
Support values for nodes are ML-bootstrap support (%)/BI-posterior probabilities.
117
Figure S3.6. Inferred G6PD gene-tree using maximum likelihood and Bayesian inference.
Support values for nodes are ML-bootstrap support (%)/BI-posterior probabilities.
119
Figure S3.7. Inferred PEPCK gene-tree using maximum likelihood and Bayesian inference.
Support values for nodes are ML-bootstrap support (%)/BI-posterior probabilities.
121
Figure S3.8. bGMYC analysis: A) Heat map of coalescent units, color indicates probability
(red = p0-0.05; dark orange = p0.05-0.5; orange = p0.5-0.9; yellow = p0.9-0.95; light
yellow = p0.95-1.0). B) Placement of individuals on tree.
123
Figure S3.9. *BEAST tree selected as the most supported species delimitation scenario
according to the path sampling/stepping-stone analyses.
125
Chapter Four:
Phylogenetic relationships of the flesh fly genus Ravinia (Diptera:
Sarcophagidae).1
1 This manuscript has been formatted for submission to Molecular Phylogenetics and Evolution
126
4.1 ABSTRACT
The genus Ravinia Robineau-Desvoidy, 1863, is a genus in the flesh fly family
Sarcophagidae and represents an agriculturally important group of insects. Many of the
species in this genus are sympatric and broadly distributed throughout North America.
Recent work conducted within this genus has identified cryptic species through the
molecular species delimitation of Ravinia anxia and Ravinia querula. In addition, previous
work using only mitochondrial loci with standard Bayesian inference and maximum
likelihood analyses, inferred Ravinia floridensis and Ravinia lherminieri to be paraphyletic
species. To investigate the phylogenetic status of the three cryptic species (R. anxia (sensu
stricto), R. querula (sensu stricto), Ravinia n.sp.), and the status of R. floridensis and R.
lherminieri, we constructed a molecular phylogeny using 12 molecular markers and
expanded sampling to include additional Ravinia species. Bayesian inference, maximum
likelihood, and coalescent-based species-tree methods using these markers strongly
supported the distinctiveness of each of the Ravinia species, including showing the three
cryptic species as monophyletic lineages. However, the paraphyletic species R. floridensis
and R. lherminieri were not resolved as monophyletic in the concatenated phylogenetic
analyses. This study is the first to employ a coalescent-based species-tree approach within
the Sarcophagidae using a large molecular data set.
Keywords: multilocus, species-tree, North America, Chaetoravinia, cryptic species
127
4.2 INTRODUCTION
Members of Ravinia Robineau-Desvoidy, 1863, are a genus of flesh flies (Diptera:
Sarcophagidae) that represent a lineage of agriculturally important insects. Primarily,
Ravinia are directly involved in the recycling of nutrients in the environment through the
breakdown of mammalian dung from a variety of livestock animals (i.e., cattle, horses, and
pigs) as a result of larval feeding. These flies are broadly distributed worldwide with a total
of 34 described species. Species diversity in this genus is greatly enriched in the New World
with 33 species found in the North and South Americas; only a single species has an Old
World distribution (Pape 1996; Pape and Dahlem 2010). The currently defined genus,
Ravinia, has a rich taxonomic history where the genus had been separated into multiple
genera or merged into a single genus at various times by various authors (Wong et al.
2015). The currently recognized genus, Ravinia (sensu lato) (Pape 1996) is composed of
three synonymized genera: Ravinia, Adinoravinia Townsend, 1917, and Chaetoravinia
Townsend, 1917 (Lopes 1969). In addition to revision at the generic level, many species of
Ravinia have proven to be taxonomically difficult groups, with several species providing
limited numbers of diagnostically relevant morphological characters. Ravinia was first
established as a genus by Robineau-Desvoidy (1863) based on the shape of the abdomen,
presence of setae on the primary segments of abdomen (female), and lack of setae on the
hind tibia (male). Since the first description of Ravinia, these characteristics are no longer
unique among the modern genera in the family Sarcophagidae (Wong et al. 2015). At
present, identification of this genus is based on the following characters: 1) postalar wall
setulate; 2) tegulae orangish in ground color; 3) frontal vitta of female at midpoint more
128
than two times the width of parafrontal plate; 4) males lacking apical scutellar bristles and
with a reddish-orange genital capsule.
The first and primary tools used in the classification of Sarcophagidae species have
been morphological characters, overwhelmingly from male reproductive features.
Typically highly trained taxonomists would perform these identifications and
classifications using diagnostic character traits unique to the species of interest. However,
the ultimate utility of morphological taxonomy in species that are cryptic or have limited
numbers of distinctive species-level traits has been noted (Hebert et al. 2004a, 2004b;
Lowenstein et al. 2009; Spillings et al. 2009; Saitoh et al. 2015). The ability to integrate
genetic data into the taxonomy of species would provide an objective and reproducible
method for evolutionary reconstructions (Díaz-Rodríguez et al. 2015). There is a short list
of studies that have used morphology in detailed evolutionary relationships within the
Sarcophaginae (Roback 1954; Lopes 1969, 1982; Pape 1996; Giroux et al. 2010). Roback
(1954) provided the first detailed phylogeny of the Sarcophaginae using morphological
characters. The most recent analysis conducted by Giroux et al. (2010) for 76 species
composing 19 genera within Sarcophaginae was based on 73 morphological characters.
However, the studies mentioned above focus on relationships above the species level and
the morphological characters used at these levels are not likely to be informative when
looking at the species level. Using genetic data can be an informative tool, especially in
systems potentially lacking morphological differentiation.
Stamper et al. (2013) explored the phylogenetic relationships of flesh flies in the
subfamily Sarcophaginae using three mitochondrial gene fragments. This study supported
the genus Oxysarcodexia (Townsend, 1917) as the sister-group to Ravinia; a relationship
129
originally proposed based on morphological evidence (Lopes 1982; Pape 1994; Giroux et
al. 2010). Wong et al. (2015) presented a phylogeny of the genus Ravinia inferred from two
mitochondrial gene fragments and noted that for some species of North American Ravinia,
discord was present between morphology used in identifications and the lineages inferred
by molecular data. The discordance could have been a result of incomplete lineage sorting
(ILS), introgression, or the presence of undescribed species. This discord was particularly
problematic in one group containing Ravinia anxia and Ravinia querula and in another
group containing Ravinia floridensis and Ravinia lherminieri. However, Wong et al. (2015)
at the time could not distinguish between these hypotheses using solely mitochondrial
data. In the most recent study involving the genus Ravinia, Wong et al. (c.f. Chapter 3)
delimited the problematic species R. anxia and R. querula using a combination of
coalescent-based species-tree analyses, migration analyses, population genetic analyses,
and phylogenetic inference. They discovered that R. anxia and R. querula, as currently
defined by morphology, are comprised of three cryptic species. Hereafter, these three
cryptic species will be referred to as R. anxia (sensu stricto=s.s.), R. querula (s.s.), and
Ravinia n.sp.
For this study we examine the molecular phylogeny of 12 species of the genus
Ravinia. We sequenced portions of five mitochondrial (mtDNA) genes and seven nuclear
(nuDNA) genes and analyzed these loci using concatenated phylogenetic (Bayesian
inference and maximum likelihood) and coalescent-based species-tree methods.
130
4.3 MATERIALS AND METHODS
Taxon selection & sampling
Specimens used for molecular analyses were collected using hand nets, Malaise
traps, or baited traps (using dog feces or aged chicken). All specimens were frozen and
taken to lab until sorted, pinned, and identified by GAD using morphological characters, or
in the case of cryptic species identified by species delimitation methods employed in Wong
et al. (c.f. Chapter 3). DNA was extracted only from legs on one side of specimen, ensuring
that all morphological structures important for species identification would be preserved.
The removed legs were kept, until DNA extraction, at -80˚C in a separate vial of 95% EtOH.
The remainder of the specimens were then pinned or stored in 95% EtOH at -80˚C and are
currently held by GAD as vouchers (will be placed in an established natural history
collection at the completion of ongoing studies).
A total of four species, from two closely related genera, were used as outgroup taxa
in this study. From the genus Oxysarcodexia which is sister to Ravinia, we used two species:
Oxysarcodexia ventricosa (Wulp, 1895); O. cingarus (Aldrich, 1916). We also included in the
analyses two species from the genus Blaesoxipha: Blaesoxipha arizona Pape, 1994; B.
cessator (Aldrich, 1916). The genus Blaesoxipha has been difficult to place within the
Sarcophaginae (Roback 1954; Lopes 1969, 1982; Pape 1994, 1996; Giroux et al. 2010;
Kutty et al. 2010), with one study placing it as sister to Ravinia (Kutty et al. 2010). As such,
we include both genera to solidify the monophyletic status of Ravinia, and the inference of
Oxysarcodexia as sister to Ravinia shown in previous studies (Lopes 1982; Pape 1994;
Giroux et al. 2010; Stamper et al. 2013).
131
Several individuals were sampled from multiple populations for 12 out of the 18
species of Ravinia in North America. The genetic analyses included 121 individuals (117
Ravinia + 4 outgroup taxa): Ravinia acerba (Walker, 1849) (1), R. anxia (s.s.) (7), Ravinia
derelicta (Walker, 1853) (25), Ravinia errabunda (Wulp, 1895) (1), Ravinia floridensis
(Aldrich, 1916) (7), Ravinia lherminieri (Robineau-Desvoidy, 1830) (8), Ravinia n.sp. (12),
Ravinia planifrons (Aldrich, 1916) (7), Ravinia pusiola (Wulp, 1895) (11), R. querula (s.s.)
(17), Ravinia stimulans (Walker, 1849) (18), Ravinia vagabunda (Wulp, 1895) (2). The
geographic distribution of sampling localities across the United States can be seen in Fig.
4.1. The following species were not included in this study because we were unable to
obtain material, even though collection efforts were conducted in regions where they are
known to occur: Ravinia anandra (Dodge, 1956), Ravinia coachellensis (Hall, 1931), Ravinia
effrenata (Walker, 1861), Ravinia pectinata (Aldrich, 1916), Ravinia sueta (Wulp, 1895),
Ravinia tancituro Roback, 1952. List of specimens included in this study, along with
GenBank accession numbers and locality information are provided in Table 4.1 and Table
S4.1.
DNA extraction, PCR, & sequencing
Qiagen’s DNeasy kits (Qiagen Cat. No. 69506) were used to extract genomic DNA
from a whole fly leg that was sliced into fragments. A total of 12 gene fragments were PCR-
amplified, including five mitochondrial (mtDNA) and seven nuclear (nuDNA) genes:
mtDNA- cytochrome oxidase subunits I and II (COI and COII, respectively), NADH
dehydrogenase 4 (ND4), NADH dehydrogenase 6 (ND6), ribosomal RNA 16S (16S); nuDNA-
elongation factor 1 alpha (EF-1a), glucose-6-phosphate dehydrogenase (G6PD), internal
transcribed spacer 2 (ITS2), phosphoenolpyruvate carboxykinase (PEPCK), period (Per),
132
ribosomal RNA 28S, white (Wt). The genes COI, COII, and ND4 were amplified using
primers pairs and specifications given in Stamper et al. (2013). A fragment of 28S spanning
the D3-D5 regions was amplified in a single fragment using previously published protocol
and primer pairs (Friedrich and Tautz 1997a, 1997b; DeBry et al. 2010). Portions of the
nuclear genes EF-1a, G6PD, PEPCK, Per, and Wt were amplified using primer pairs and
protocols given in Wong et al. (c.f. Chapter 3). For the ITS2 amplification, we used primer
pairs (without tailed-ends) and protocol given in Song et al. (2008b). For the 16S gene, we
used primer pairs (without tailed-ends) and protocols given in Kutty et al. (2007). All
amplifications were conducted in a solution consisting of: 200 μM dNTPs (Promega), 1μM
of each primer, 1.5mM MgCl2, 0.25 units Platinum Pfx DNA polymerase (Invitrogen), buffer
(final concentration: 50mM Tris, pH 8.5, 4 mM MgCl2, 20 mM KCl, 500 μg/ml bovine serum
albumin, 5% DMSO) (Idaho Technology), and 3μl template DNA.
Forward and reverse DNA strands were sequenced with their respective primer on
an ABI prism 3730x/ with base calling and data compilation by Beckman Coulter Genomics,
Inc. (Danvers, MA, USA).
Data alignment, assembly, & characterization
Sequence chromatograms were visually checked in FinchTV v.1.4.0 (Geospiza, Inc.
www.geospiza.com) for heterozygous sites and potential sequencing errors. After general
quality control, sequences were assembled and edited in CLC Main Workbench v.7.0.3 (CLC
Bio www.clcbio.com). All assembled sequences were aligned in Mesquite v.2.73 (Maddison
and Maddison 2011). The program MARNA v.3.4.2 (Siebert and Backofen 2005) was used
for the preliminary alignments of 16S and 28S taking into account secondary structure of
the rRNA sequence. For all other loci, the program Muscle v.3.8 (Edgar 2004) was used for
133
the sequence alignments. The resulting alignments for the protein coding loci were re-
checked as both the translated amino acid sequence and the nucleotide sequence. No
insertions or deletions (indels) were present among sequences for the following genes:
COI, COII, ND4, ND6, 16S, 28S, EF-1a, G6PD, PEPCK, Wt. As such, the final alignments for
these genes were the same as preliminary Muscle and MARNA alignments. Indels were
found to be present within both ITS2 and Per. For ITS2 there were several indel differences
between species, however within species the presence and location of the indels were the
same. The Per alignment had two indel locations on the portion of the gene sequenced,
however these did not cause any change in amino acid reading-frame and did not produce
premature stop codons. Based on the examination of sequence characteristics (i.e., quality
score (phred) of chromatograms, amplicon length; Song et al. 2008a), absence of in-frame
stop codons, and use of taxon specific primers, we believe that silenced genes and nuclear
insertions of mitochondrial DNA (numts) were not erroneously sequenced.
Nuclear gene sequences were tested for recombination with the program TOPALi
v.2.5 (Milne et al. 2009) using default settings and the Difference of Sums of Squares (DSS)
method. General sequence characterization was performed using the program DNasp v.5
(Librado and Rozas 2009); evaluating the number of parsimony informative sites (PI),
segregating sites (S), and nucleotide diversity (π). The gene information characterized was
for the data set including all Ravinia sequences (i.e., outgroup sequences were excluded).
Concatenated multi-locus analyses
The gene partition scheme and model of sequence evolution were selected using the
program PartitionFinder (Lanfear et al. 2012). The Akaike Information Criterion (AIC;
Akaike 1973, 1974) was used to compare candidate models of sequence evolution and
134
partition scheme. Only those model choice/partition schemes that could be implemented in
the phylogenetic programs (RAxML and MrBayes) were considered in PartitionFinder
analyses using the functions “models=RAxML” (for maximum likelihood) or
“models=mrbayes” (for Bayesian inference). We analyzed the concatenated mitochondrial
and nuclear loci with maximum likelihood (ML) using RAxML v.8.0.20 (Stamatakis 2014),
and Bayesian inference (BI) using the program MrBayes v.3.2.1 (Ronquist and Huelsenbeck
2003).
ML analyses of the concatenated data set, conducted using RAxML, used subsets of
the GTRGAMMA model (Appendix S4.1) selected by PartitionFinder. The RAxML program
defaults were used to stop search and nodal support was assessed by bootstrap resampling
(Felsenstein 1985) with 5000 pseudoreplicates. Maximum likelihood bootstrap
probabilities (BPML) were mapped onto the 70% majority rule consensus tree of replicates
using the program SumTrees implemented in DendroPy v.3.12.0 (Sukumaran and Holder
2010). BPML ≥ 95% are considered high support in these analyses.
BI analyses of the concatenated data set, conducted with MrBayes, used two
independent Metropolis-coupled Markov chain Monte Carlo (MC3) (Geyer 1991; Gilks and
Roberts 1996) analyses to sample from the posterior distribution. Each MC3 analysis had
one cold and three heated chains with default settings, sampling 1000 generations for
2x107 generations and discarding the first 50% of samples as burn-in. The program
TRACER v.1.6 (Rambaut et al. 2014) was used to assess run convergence by examining: the
harmonic mean, stability of parameter values, similarity of branch lengths, trees, and
likelihood values of chains, and the value of average standard deviation of split frequencies.
135
Nodal support was based on Bayesian posterior probabilities (PP). Nodes with PP ≥ 0.95
are considered to be high support.
Species-tree analysis
The program *BEAST (Heled and Drummond 2010) implemented in the program
BEAST (Drummond et al. 2012), was used to estimate the species-tree under the
multispecies coalescent model. *BEAST is able to estimate the species-tree directly from
the sequence data and calculates the posterior probability for gene-trees, species-tree, and
divergence times. To make calculations tractable, a priori information must be used to
assign individuals to species groups in the species-tree. Species groups were assigned
based on morphological identifications for all species, except for R. anxia (s.s.), R. querula
(s.s.), and Ravinia n.sp., which were identified based on the molecular species delimitation
of Wong et al. (c.f. Chapter 3). Two data sets were used in *BEAST analyses to reconstruct
the species-tree (nuDNA, nuDNA + mtDNA). The full data set consisted of 12 loci (5 mtDNA
+ 7 nuDNA) and 16 species (including 4 outgroup taxa). Since mtDNA is maternally
inherited as a single unit, all mtDNA loci will have the same gene-tree. As such, the 5
mitochondrial loci were analyzed as a linked unit in the tree for *BEAST analyses. *BEAST
analyses used an uncorrelated lognormal relaxed clock (Drummond et al. 2006) with a Yule
model prior of divergence times. The MCMC algorithm was run for 2x107 generations,
sampling every 1000 generations and discarding the first 50% as burn-in. Convergence of
runs were assessed by checking for stationary in tree lengths and likelihood scores using
Tracer v.1.6 looking at stationary of branch lengths and likelihood scores. PartitionFinder
was used to select the best-fitting model of nucleotide substitution from among those
implemented in *BEAST. The program TreeAnnotator v.1.8.0 (Heled and Drummond 2010)
136
was used to summarize the species-trees into a single maximum clade credibility tree, and
visualized using FigTree v.1.4.1 (http://tree.bio.ed.ac.uk/software/figtree/). In addition,
DensiTree (Bouckaert 2010) was used to visualize the entire posterior distribution of
species-trees. This program takes all 1x104 species-trees, each represented as a single
semi-transparent line, and overlays them on top of each other. The areas where there is
much agreement in topology and branch lengths results in many lines being drawn, giving a
darker color.
4.4 RESULTS
Molecular data
DNA sequence data were obtained for 121 individuals (including 4 outgroup
specimens) comprising 12 of the 18 species of Ravinia that are found in North America
(Pape 1996; Pape and Dahlem 2010; Wong et al. 2015; c.f. Chapter 3). A total of 8188 base
pairs (bp) of DNA sequence data (mtDNA = 4399 bp; nuDNA= 3789 bp) were included in
the final molecular alignment. The list of samples used in molecular analyses, locality
information, and GenBank accession numbers augmenting those published in (Stamper et
al. 2013; Wong et al. 2015; c.f. Chapter 3) are provided in Table 4.1 and Table S4.1.
Information for sequenced loci, including nucleotide diversity, are given in Table 4.2.
Concatenated multi-locus analysis
In general, the concatenated phylogenetic analyses using both ML and BI inferred
the same topology with no significant disagreement between the two analyses. As such,
only the BI tree is shown in Fig. 4.2. The only disagreement between ML and BI trees was
that the ML analyses could not resolve the R. acerba/R. pusiola clade and the BI analyses
did, albeit with low nodal support (Fig. 4.2; PP=0.57). Six out of the twelve Ravinia species
137
included in this study were resolved with high support, allowing differentiation between
these closely related lineages. Those species not able to be resolved as monophyletic with
high nodal support are R. acerba, R. lherminieri, and R. derelicta.
Both ML and BI analyses found strong support for two clades within the modern
Ravinia (sensu lato) that correspond to previously synonymized genera, Chaetoravinia
Townsend, 1917 and Ravinia (Fig 4.2; BPML=100; PP=1.0). These groups are also supported
by a distinct morphological character: absence or presence of setae on the R1 wing vein in
Ravinia and Chaetoravinia, respectively (Fig. 4.3; Lopes 1969). For additional information
on the previously synonymized genera, Chaetoravinia and Ravinia, the authors refer
readers to Wong et al. (2015). Within the Ravinia group, R. planifrons is found to be sister to
the other seven species: R. anxia (s.s.), R. floridensis, R. lherminieri, Ravinia n.sp., R. pusiola,
R. querula (s.s.) (Fig. 4.2; BPML= 100; PP= 1.0). Our seven specimens of R. planifrons are
monophyletic and distinct from the other species (Fig. 4.2; BPML= 100; PP= 1.0). Ravinia
pusiola specimens sorted onto two distinct haplotype groups: a paraphyletic group
comprised of our single specimen of R. acerba plus two R. pusiola specimens, which was
sister to a clade of the remaining R. pusiola (Fig. 4.2; BPML= 100; PP=1.0).Together R.
pusiola and R. acerba form a group that is sister to a clade consisting of R. anxia (s.s.), R.
floridensis, R. lherminieri, Ravinia n.sp., and R. querula (s.s.) (Fig. 4.2; BPML=100; PP= 1.0).
Our eight specimens of R. lherminieri sorted onto two distinct haplotypes: the R. lherminieri
specimens from New York form a highly supported paraphyletic group with all the R.
floridensis specimens (Fig. 4.2; BPML= 100; PP= 1.0), which was sister to a highly supported
clade of R. lherminieri from Florida and Tennessee (Fig. 4.2; BPML= 100; PP= 1.0). In the
group comprising R. anxia (s.s.), R. querula (s.s.), and Ravinia n.sp., we found highly
138
supported reciprocal monophyly between these three species. Our eight specimens of R.
anxia (s.s.) formed a highly supported clade (Fig. 4.2; BPML= 100; PP= 1.0) that was sister to
a clade comprised of R. querula (s.s.) and Ravinia n.sp. (Fig. 4.2; BPML=100; PP=1.0). Both R.
querula (s.s.) and Ravinia n.sp. were resolved as monophyletic groups with high nodal
support (Fig. 4.2; BPML= 100; PP=1.0).
Within the Chaetoravinia group, R. derelicta, R. stimulans, and R. vagabunda formed a
clade with high nodal support, which was sister to our single specimen of R. errabunda (Fig.
4.2; BPML= 100; PP= 1.0). Ravinia stimulans was resolved as a monophyletic group and was
found to be sister to a clade comprising R. derelicta + R. vagabunda with high nodal support
(Fig. 4.2; BPML= 100; PP= 1.0). Our two R. vagabunda individuals, collected from a single
locality in New Mexico, were monophyletic and strongly supported as nested within a
paraphyletic R. derelicta (Fig. 4.2; BPML= 100; PP= 1.0).
Species-tree analysis
The species-trees of the mitochondrial and nuclear data set were estimated using
*BEAST, to evaluate the inferred phylogeny of the concatenated data set. The species-trees
were summarized in a maximum clade credibility tree (MCCT; Fig. 4.4A) and a cloudogram
(Fig. 4.4B). The species-trees showed similar results when compared with the
concatenated multi-locus analyses. The species-trees strongly supported grouping R.
lherminieri and R. floridensis with a posterior probability of 1.0 (Fig. 4.4A). In addition, the
species-trees grouped R. acerba and R. pusiola with high nodal support (Fig. 4.4A; PP= 1.0),
and grouped R. vagabunda and R. derelicta with high nodal support (Fig. 4.4A; PP= 1.0).
Due to our low sample sizes for R. acerba and R. vagabunda, the disagreement between the
139
concatenated phylogeny and the species-tree are unclear with regard to the paraphyletic
relationships.
4.5 DISCUSSION
The present study uses multi-locus analyses on a concatenated data set and
coalescent-based species-tree estimations to infer relationships within the genus, Ravinia.
According to previous studies, several species of Ravinia were inferred as paraphyletic
using two mitochondrial loci (Wong et al. 2015). Although mitochondrial loci (e.g., COI +
COII) are often recognized as markers displaying high levels of polymorphic divergence
(i.e., barcoding loci), COI + COII sequences in some cases may be unsuitable in
differentiating phylogenetic relationships among closely related species (Nelson et al.
2007; Xiao et al. 2010). Within Ravinia there are several clades that contain
morphologically distinct species that are for the most part not distinguishable using COI +
COII sequence data (e.g., R. floridensis/R. lherminieri, R. acerba/R. pusiola, and R.
derelicta/R. vagabunda; Wong et al. 2015). In this study, we used five mitochondrial loci
and seven nuclear loci with increased taxonomic sampling over previous studies (Wong et
al. 2015; c.f. Chapter 3).
Based on the phylogenetic tree (Fig. 4.2), the species of Ravinia fall into two distinct
clades (sections A and B), with high nodal support (BPML= 100; PP= 1.0). These clades
represent the groups Ravinia and Chaetoravinia, previously synonymized into the currently
recognized genus Ravinia (sensu lato). The separation of these groups is congruent with a
previous study using only mitochondrial data to infer species relationships (Wong et al.
2015). This separation is further supported morphologically by the absence or presence of
setae on the R1 wing vein (Fig 4.3) in Ravinia and Chaetoravinia, respectively. This provides
140
further evidence for recognition of these two groups as subgenera of Ravinia (sensu lato) or
even as separate genera. Further sampling of individuals thought to be part of the
Adinoravinia lineage would be needed to solidify the status of these groups.
There was a single difference between the topology inferred from the concatenated
data set in this study and the mtDNA phylogeny presented in Wong et al. (2015). Wong et
al. (2015) showed that R. planifrons formed a highly supported monophyletic clade that is
sister to a clade comprised of the species: R. anxia (s.s.), R. floridensis, R. lherminieri, Ravinia
n.sp., and R. querula (s.s.). In this study, however, when the species R. acerba is added to
analyses along with additional loci, the clade R. acerba + R. pusiola is found to be sister to
the clade comprised of: R. anxia (s.s.), R. floridensis, R. lherminieri, Ravinia n.sp., and R.
querula (s.s.); with high nodal support (Fig. 4.2; BPML= 100; PP= 1.0). While in both studies
the topologies were highly supported, this study presents a higher nodal support for this
relationship while including a greater amount of molecular data. However, due to the
limited number of R. acerba specimens sampled, the exact placement of this species within
Ravinia is uncertain.
Ravinia anxia (s.s.), Ravinia querula (s.s.), and Ravinia n.sp.
These three cryptic species are inferred as genetically divergent, reciprocally
monophyletic lineages with high nodal support (Fig. 4.2; BPML= 100; PP= 1.0). The
phylogeny inferred using the full concatenated data set presents a topology similar to
previous studies (Wong et al. 2015; c.f. Chapter 3). Within this clade R. anxia (s.s.) is sister
to the larger group composed of the species R. querula (s.s.) and Ravinia n.sp. (Fig. 4.2;
BPML= 100; PP= 1.0). The relationship between R. querula (s.s.) and Ravinia n.sp. is largely
congruent with geography, where these sister taxa are geographically proximate and
141
overlap in the northwest United States (Fig. 4.1). Although R. querula (s.s.) appears to have
a much larger distribution across the United States, our collected species of R. querula (s.s.)
and Ravinia n.sp. are genetically divergent in locations where the two species overlap. The
species-tree inferred using *BEAST, recovered strong support for the branching of R. anxia
(s.s.) (Fig. 4.4; PP=1.0), with relatively weak support for its sister clade comprised of
R. querula (s.s.) and Ravinia n.sp. (Fig. 4.4A; PP=0.73). The next probable branching, based
on the posterior distribution of species-trees in the cloudogram (Fig. 4.4B), has
Ravinia n.sp. and R. anxia (s.s.) as sister taxa.
Ravinia floridensis and Ravinia lherminieri
Our results using the concatenated data set show that R. floridensis and R.
lherminieri form a strongly supported clade (Fig. 4.2; BPML= 100; PP= 1.0). Within this clade
are two distinct and highly supported clades, but these are not indicative of the two species
diagnoses based on morphological identifications. One clade comprises several R.
lherminieri specimens collected from Tennessee and New York. The other clade is
comprised of all the R. floridensis individuals as well as R. lherminieri collected from New
York. This discrepancy between the morphology and the genetic lineages has also been
found in a previous study (Wong et al. 2015). Wong et al. (2015) presented several
hypotheses of how this discrepancy could have arisen between the morphology and genetic
lineages. In summary, the first possibility is that current species descriptions based on
morphology could be incorrect and the species division presented by the genetic data could
be the actual species boundaries. Second, the morphological species descriptions could be
correct, in which the discrepancy could be from ILS or hybridization. Wong et al. (2015)
state that the hybridization hypothesis is unlikely based on the large geographic separation
142
between R. floridensis specimens and those identified as R. lherminieri (Fig 4.1). Finally, the
morphological characters used in the identification of R. floridensis exhibit variation across
the species range, and as such may be indicative of cryptic species. Wong et al. (2015)
concluded that future sampling of specimens and additional molecular markers would be
needed to infer which of these scenarios is most likely. In the present study, the phylogeny
inferred using a large concatenated data set found a similar topology as Wong et al. (2015)
indicating paraphyletic relationships in R. lherminieri. When multilocus coalescent-based
analyses are performed, to compare the inferred species-tree to the concatenated RAxML
and MrBayes phylogeny, a similar relationship is highly supported. The species-tree
recovered strong support for the grouping of R. floridensis and R. lherminieri (Fig. 4.4A;
PP=1.0). Furthermore, the strong support for this grouping was found when using either
the nuDNA only or the mtDNA + nuDNA data sets (Fig. 4.4A; Fig. S4.1). To further explore
the possible causes of this discordance in R. lherminieri and R. floridensis, it is
recommended that a full species delimitation study be conducted using methods that do
not require the a priori partitioning of individuals to species categories.
4.6 CONCLUSIONS
The purpose of this study was to explore the phylogenetic relationships within
Ravinia using increased sampling of taxa and molecular loci. In addition, we included the
updated species taxonomy that has been previously suggested concerning the R. anxia and
R. querula species complex (c.f. Chapter 3). This study is the first to apply species-tree
coalescent-based analyses to species in the family Sarcophagidae. As a result of these
analyses, we can identify a case where sole use of mtDNA data would lead to mistaken
identity of species, particularly R. floridensis and R. lherminieri. All analyses found strong
143
support for the division of Ravinia into the previously recognized groups Ravinia and
Chaetoravinia. If these divisions are maintained when including specimens believed to be
part of the Adinoravinia lineage, then a formal recognition of these lineages would be
necessary. This study is only focused on North American Ravinia, but there are also
numerous other species found only in the Central and South Americas. For some species
our sampling was limited, even though there were focused collection trips throughout the
U.S. in areas where certain species are known to be encountered (Fig. 4.1). This is a
problem often encountered in molecular studies because of an inherent rarity of species
(Lim et al. 2012). Regardless of this problem several relationships were able to be resolved,
including the discordant relationships seen in previous studies (Wong et al. 2015). The
present study, however, cannot resolve those relationships where specimen sampling
appears to be a problem (e.g., R. vagabunda, R. acerba). Further specimen collections
should focus on increasing geographic sampling of species that exhibit discordance. Future
research would benefit from classical taxonomy providing new character descriptions that
could be used in morphological analyses and examining the Ravinia phylogeny with the
addition of Old World and Neotropical taxa.
144
4.7
TABL
ES
Tabl
e 4.
1. L
ist o
f spe
cim
ens,
spec
imen
iden
tific
atio
n (I
.D.)
code
s, se
quen
ced
gene
s, an
d Ge
nBan
k Ac
cess
ion
num
bers
. Sp
ecie
s I.D
. CO
I CO
II
ND
4 N
D6
16S
28S
EF-1
a G6
PD
ITS2
PE
PCK
Pe
riod
W
hite
Bl
aeso
xiph
a ar
izon
a AW
32
JQ80
6809
JQ
8067
88
JQ80
6767
KR
6769
78
------
------
- ---
------
----
------
------
- ---
------
---
------
------
- ---
------
----
------
------
---
------
---
Blae
soxi
pha
cess
ator
AW
27
JQ80
6810
JQ
8067
89
JQ80
6768
KR
6769
79
------
------
- ---
------
----
KR60
8322
KR
6767
59
------
------
- KR
6767
06
KR67
6812
KR
6766
30
Oxys
arco
dexi
a ci
ngar
us
AP68
JQ
8068
16
JQ80
6795
JQ
8067
74
------
------
- ---
------
----
------
------
- ---
------
----
------
------
---
------
----
------
------
- ---
------
---
------
------
Ox
ysar
code
xia
vent
rico
sa
E08
GQ22
3321
GQ
2233
21
GQ22
3292
KR
6769
80
------
------
- ---
------
----
KR60
8323
KR
6767
60
------
------
- KR
6767
07
KR67
6813
KR
6766
31
Ravi
nia
acer
ba
AU02
KR
6083
06
KR67
7020
---
------
----
------
------
- KR
6769
44
KR67
6861
---
------
----
------
------
KR
6769
90
------
------
- ---
------
---
------
------
Ravi
nia
n.sp
. AE
79
KR60
8307
KR
6770
21
------
------
- ---
------
----
------
------
- ---
------
----
KR60
8324
KR
6767
61
------
------
- KR
6767
08
KR67
6814
KR
6766
32
Rav
inia
n.sp
. AE
80
KR60
8308
KR
6770
22
------
------
- ---
------
----
------
------
- ---
------
----
KR60
8325
KR
6767
62
------
------
- KR
6767
09
KR67
6815
KR
6766
33
Rav
inia
n.sp
. AE
81
JQ80
7069
KJ
5864
03
KR67
6988
KR
6769
81
------
------
- KR
6768
62
KR60
8326
KR
6767
63
------
------
- KR
6767
10
KR67
6816
KR
6766
34
Rav
inia
n.sp
. AF
03
KJ58
6379
KJ
5864
66
------
------
- ---
------
----
------
------
- KR
6769
15
KR60
8372
KR
6767
88
------
------
- KR
6767
35
KR67
6838
KR
6766
76
Rav
inia
n.sp
. AG
31
KJ58
6383
KJ
5864
70
------
------
- ---
------
----
------
------
- KR
6769
19
KR60
8376
KR
6767
92
------
------
- KR
6767
39
KR67
6842
KR
6766
80
Rav
inia
n.sp
. AG
51
KJ58
6343
KJ
5864
04
------
------
- ---
------
----
------
------
- KR
6768
63
KR60
8327
---
------
---
------
------
- ---
------
----
------
------
KR
6766
35
Rav
inia
n.sp
. AH
52
KJ58
6344
KJ
5864
05
------
------
- ---
------
----
KR67
6945
KR
6768
64
KR60
8328
KR
6767
64
KR76
991
KR67
6711
KR
6768
17
KR67
6636
Rav
inia
n.sp
. AJ
18
KJ58
6345
KJ
5864
06
------
------
- ---
------
----
------
------
- KR
6768
65
KR60
8329
KR
6767
65
------
------
- KR
6767
12
KR67
6818
KR
6766
37
Rav
inia
n.sp
. AJ
19
KJ58
6388
KJ
5864
75
------
------
- ---
------
----
------
------
- KR
6769
24
KR60
8381
KR
6767
97
------
------
- KR
6767
44
KR67
6847
KR
6766
85
Rav
inia
n.sp
. AK
04
KR60
8316
KR
6770
28
------
------
- ---
------
----
------
------
- ---
------
----
KR60
8383
KR
6767
99
------
------
- KR
6767
46
KR67
6849
KR
6766
87
Rav
inia
n.sp
. AK
25
KJ58
6346
KJ
5864
07
------
------
- ---
------
----
------
------
- KR
6768
66
KR60
8330
KR
6767
66
------
------
- KR
6767
13
KR67
6819
KR
6766
38
Rav
inia
n.sp
. BA
16
JQ80
7102
KJ
5864
81
------
------
- ---
------
----
------
------
- KR
6769
30
KR60
8389
KR
6768
05
------
------
- KR
6767
52
KR67
6855
KR
6766
93
Ravi
nia
anxi
a (s
.s.)
AT44
KJ
5863
47
KJ58
6408
---
------
----
------
------
- ---
------
----
KR67
6867
---
------
----
------
------
---
------
----
------
------
- ---
------
---
------
------
R. a
nxia
(s.s.
) AT
45
JQ80
7068
KJ
5864
09
KR67
6989
KR
6769
82
------
------
- ---
------
----
------
------
- ---
------
---
------
------
- ---
------
----
------
------
---
------
---
R. a
nxia
(s.s.
) AT
46
KJ58
6348
KJ
5864
10
------
------
- ---
------
----
------
------
- ---
------
----
------
------
- ---
------
---
------
------
- ---
------
----
------
------
---
------
---
R. a
nxia
(s.s.
) AT
47
KJ58
6349
KJ
5864
11
------
------
- ---
------
----
------
------
- ---
------
----
------
------
- ---
------
---
------
------
- ---
------
----
------
------
---
------
---
R. a
nxia
(s.s.
) AT
49
KJ58
6350
KJ
5864
12
------
------
- ---
------
----
------
------
- ---
------
----
------
------
- ---
------
---
------
------
- ---
------
----
------
------
---
------
---
R. a
nxia
(s.s.
) AT
50
JQ80
7070
KJ
5864
13
------
------
- ---
------
----
------
------
- ---
------
----
------
------
- ---
------
---
------
------
- ---
------
----
------
------
---
------
---
145
R. a
nxia
(s.s.
) AW
48
KJ58
6351
KJ
5864
14
------
------
- ---
------
----
------
------
- ---
------
----
KR60
8331
KR
6767
67
------
------
- KR
6767
14
KR67
6820
KR
6766
39
R. a
nxia
(s.s.
) AW
70
KR60
8309
KR
6770
23
------
------
- ---
------
----
------
------
- ---
------
----
KR60
8332
KR
6767
68
------
------
- KR
6767
15
KR67
6821
KR
6766
40
Ravi
nia
dere
licta
AA
10
KJ58
6352
KJ
5864
15
------
------
- ---
------
----
KR67
6946
---
------
----
------
------
- ---
------
---
KR67
6992
---
------
----
------
------
KR
6766
41
R. d
erel
icta
AA
13
KJ58
6353
KJ
5864
16
------
------
- ---
------
----
KR67
6947
---
------
----
------
------
- ---
------
---
KR67
6993
---
------
----
------
------
KR
6766
42
R. d
erel
icta
AA
25
KJ58
6354
KJ
5864
17
------
------
- ---
------
----
KR67
6948
---
------
----
KR60
8333
---
------
---
KR67
6994
---
------
----
------
------
KR
6766
43
R. d
erel
icta
AA
26
KJ58
6355
KJ
5864
18
------
------
- ---
------
----
KR67
6949
KR
6768
68
------
------
- ---
------
---
KR67
6995
---
------
----
------
------
---
------
---
R. d
erel
icta
AA
38
KJ58
6356
KJ
5864
19
------
------
- ---
------
----
KR67
6950
KR
6768
69
------
------
- ---
------
---
KR67
6996
---
------
----
------
------
---
------
---
R. d
erel
icta
AB
05
KJ58
6357
KJ
5864
20
------
------
- ---
------
----
KR67
6951
KR
6768
70
------
------
- ---
------
---
KR67
6997
---
------
----
------
------
---
------
---
R. d
erel
icta
AB
41
KJ58
6358
KJ
4864
21
------
------
- ---
------
----
KR67
6952
KR
6768
71
------
------
- ---
------
---
KR67
6998
---
------
----
------
------
KR
6766
44
R. d
erel
icta
AC
75
KJ58
6359
KJ
5864
22
------
------
- ---
------
----
KR67
6953
KR
6768
72
------
------
- ---
------
---
KR67
6999
---
------
----
------
------
KR
6766
45
R. d
erel
icta
AC
76
KJ58
6360
KJ
5864
23
------
------
- ---
------
----
KR67
6954
---
------
----
------
------
- ---
------
---
KR67
7000
---
------
----
------
------
---
------
---
R. d
erel
icta
AC
81
KJ58
6361
KJ
5864
24
------
------
- ---
------
----
KR67
6955
KR
6768
73
KR60
8334
---
------
---
KR67
7001
---
------
----
------
------
---
------
---
R. d
erel
icta
AD
01
KJ58
6362
KJ
5864
25
------
------
- ---
------
----
KR67
6956
KR
6768
74
KR60
8335
---
------
---
KR67
7002
---
------
----
------
------
---
------
---
R. d
erel
icta
AD
02
KJ58
6363
KJ
5864
26
------
------
- ---
------
----
KR67
6957
KR
6768
75
KR60
8336
KR
6767
69
KR67
7003
KR
6767
16
------
------
KR
6766
46
R. d
erel
icta
AD
06
KJ58
6364
KJ
5864
27
------
------
- ---
------
----
KR67
6958
KR
6768
76
KR60
8337
---
------
---
KR67
7004
---
------
----
------
------
---
------
---
R. d
erel
icta
AD
07
KJ58
6365
KJ
5864
28
------
------
- ---
------
----
KR67
6959
KR
6768
77
KR60
8338
---
------
---
KR67
7005
---
------
----
------
------
---
------
---
R. d
erel
icta
AD
11
KJ58
6366
KJ
5864
29
------
------
- ---
------
----
KR67
6960
KR
6768
78
------
------
- ---
------
---
KR67
7006
---
------
----
------
------
---
------
---
R. d
erel
icta
AD
40
KJ58
6367
KJ
5864
30
------
------
- ---
------
----
KR67
6961
KR
6768
79
KR60
8339
---
------
---
KR67
7007
---
------
----
------
------
---
------
---
R. d
erel
icta
AD
42
KJ58
6368
KJ
5864
31
------
------
- ---
------
----
KR67
6962
KR
6768
80
KR60
8340
---
------
---
KR67
7008
---
------
----
------
------
---
------
---
R. d
erel
icta
AE
57
KJ58
6369
KJ
5864
32
------
------
- ---
------
----
------
------
- KR
6768
81
KR60
8341
KR
6767
70
------
------
- KR
6767
17
KR67
6822
KR
6766
47
R. d
erel
icta
AN
13
KJ58
6370
KJ
5864
33
------
------
- ---
------
----
KR67
6963
KR
6768
82
KR60
8342
---
------
---
KR67
7009
---
------
----
------
------
---
------
---
R. d
erel
icta
AN
18
KJ58
6371
KJ
5864
34
------
------
- ---
------
----
------
------
- KR
6768
83
KR60
8343
---
------
---
------
------
- ---
------
----
------
------
---
------
---
R. d
erel
icta
AT
48
KR60
8310
KR
6770
24
------
------
- ---
------
----
------
------
- KR
6768
84
------
------
- ---
------
---
------
------
- ---
------
----
------
------
---
------
---
R. d
erel
icta
AZ
09
JQ80
7072
KJ
5864
35
------
------
- ---
------
----
------
------
- ---
------
----
------
------
- ---
------
---
------
------
- ---
------
----
------
------
KR
6766
48
R. d
erel
icta
AZ
62
JQ80
7073
KJ
5864
36
------
------
- ---
------
----
------
------
- ---
------
----
------
------
- ---
------
---
------
------
- ---
------
----
------
------
---
------
---
R. d
erel
icta
BA
29
JQ80
7075
KJ
5864
37
------
------
- ---
------
----
KR67
6964
KR
6768
85
KR60
8344
---
------
---
KR67
7010
---
------
----
------
------
---
------
---
R. d
erel
icta
BA
30
JQ80
6817
JQ
8067
96
JQ80
6775
KR
6769
83
------
------
- KR
6768
86
------
------
- ---
------
---
------
------
- ---
------
----
------
------
---
------
---
R. e
rrab
unda
AE
51
KJ58
6372
KJ
5864
38
------
------
- ---
------
----
KR67
6965
KR
6768
87
KR60
8345
KR
6767
71
KR67
011
KR67
6718
KR
6768
23
KR67
6649
146
Ravi
nia
flori
dens
is AA
01
JQ80
7076
KJ
5864
39
------
------
- ---
------
----
KR67
6966
KR
6768
88
KR60
8346
KR
6767
72
------
------
- KR
6767
19
KR67
6824
KR
6766
50
R. f
lori
dens
is AA
03
JQ80
7077
KJ
5864
40
------
------
- ---
------
----
KR67
6967
---
------
----
KR60
8347
KR
6767
73
------
------
- KR
6767
20
KR67
6825
KR
6766
51
R. f
lori
dens
is AA
04
KJ58
6373
KJ
5864
41
------
------
- ---
------
----
KR67
6968
KR
6768
89
KR60
8348
KR
6767
74
KR67
7012
KR
6767
21
KR67
6826
KR
6766
52
R. f
lori
dens
is AA
60
JQ80
7078
KJ
5864
42
------
------
- ---
------
----
KR67
6969
KR
6768
90
KR60
8349
KR
6767
75
KR67
7013
KR
6767
22
KR67
6827
KR
6766
53
R. f
lori
dens
is AB
15
JQ80
7079
KJ
5864
43
------
------
- ---
------
----
KR67
6970
KR
6768
91
KR60
8350
KR
6767
76
KR67
7014
KR
6767
23
KR67
6828
KR
6766
54
R. f
lori
dens
is AD
31
KR60
8311
KR
6770
25
------
------
- ---
------
----
------
------
- ---
------
----
KR60
8351
KR
6767
77
------
------
- KR
6767
24
KR67
6829
KR
6766
55
R. f
lori
dens
is AD
41
KJ58
6374
KJ
5864
44
------
------
- ---
------
----
KR67
6971
KR
6768
92
KR60
8352
KR
6767
78
KR67
7015
KR
6767
25
KR67
6830
KR
6766
56
Ravi
nia
lher
min
ieri
AA
36
JQ80
7080
KJ
5864
45
------
------
- ---
------
----
KR67
6972
KR
6768
93
KR60
8353
KR
6767
79
KR67
7016
KR
6767
26
------
------
KR
6766
57
R. l
herm
inie
ri
AN14
JQ
8070
81
KJ58
6446
---
------
----
------
------
- ---
------
----
KR67
6894
KR
6083
54
KR67
6780
---
------
----
KR67
6727
KR
6768
31
KR67
6658
R. l
herm
inie
ri
AN15
JQ
8070
82
KJ58
6447
---
------
----
------
------
- ---
------
----
KR67
6895
KR
6083
55
KR67
6781
---
------
----
KR67
6728
KR
6768
32
KR67
6659
R. l
herm
inie
ri
AN16
JQ
8078
18
JQ80
6797
JQ
8067
76
KR67
6984
---
------
----
KR67
6896
KR
6083
56
KR67
6782
---
------
----
KR67
6729
KR
6768
33
KR67
6660
R. l
herm
inie
ri
AN17
JQ
8070
83
KJ58
6448
---
------
----
------
------
- ---
------
----
KR67
6897
KR
6083
57
KR67
6783
---
------
----
KR67
6730
KR
6768
34
KR67
6661
R. l
herm
inie
ri
AQ58
JQ
8070
99
KJ58
6449
---
------
----
------
------
- ---
------
----
KR67
6898
---
------
----
------
------
---
------
----
KR67
6731
---
------
---
KR67
6662
R. l
herm
inie
ri
AZ44
JQ
8070
84
KJ58
6451
---
------
----
------
------
- ---
------
----
KR67
6899
KR
6083
58
KR67
6784
---
------
----
------
------
- KR
6768
35
KR67
6663
R. l
herm
inie
ri
AZ58
JQ
8070
85
KJ58
6452
---
------
----
------
------
- ---
------
----
KR67
6900
---
------
----
KR67
6785
---
------
----
------
------
- ---
------
---
KR67
6664
Ravi
nia
plan
ifron
s AE
45
KR60
8312
---
------
----
------
------
- ---
------
----
------
------
- KR
6769
01
------
------
- ---
------
---
------
------
- KR
6767
32
------
------
---
------
---
R. p
lani
fron
s AF
05
KR60
8313
---
------
----
------
------
- ---
------
----
------
------
- KR
6769
02
------
------
- ---
------
---
------
------
- ---
------
----
------
------
---
------
---
R. p
lani
fron
s AF
06
JQ80
7088
KJ
5864
53
------
------
- ---
------
----
------
------
- KR
6769
03
KR60
8359
---
------
---
------
------
- ---
------
----
------
------
KR
6766
65
R. p
lani
fron
s AG
34
KJ58
6376
KJ
5864
54
------
------
- ---
------
----
------
------
- KR
6769
04
KR60
8360
---
------
---
------
------
- ---
------
----
------
------
KR
6766
66
R. p
lani
fron
s AG
49
KJ58
6377
KJ
5864
55
------
------
- ---
------
----
------
------
- KR
6769
05
KR60
8361
KR
6767
86
------
------
- KR
6767
33
KR67
6836
KR
6766
67
R. p
lani
fron
s AK
08
JQ80
7089
KJ
5864
56
------
------
- ---
------
----
KR67
6973
KR
6769
06
KR60
8362
---
------
---
KR67
7017
---
------
----
------
------
---
------
---
R. p
lani
fron
s AK
09
KR60
8314
KR
6770
26
------
------
- ---
------
----
------
------
- ---
------
----
KR60
8363
---
------
---
------
------
- ---
------
----
------
------
KR
6766
68
Ravi
nia
pusio
la
AW64
JQ
8070
94
KJ58
6457
---
------
----
------
------
- ---
------
----
KR67
6907
KR
6083
64
------
------
---
------
----
------
------
- ---
------
---
KR67
6669
R. p
usio
la
AW73
JQ
8070
95
KJ58
6458
---
------
----
------
------
- ---
------
----
KR67
6908
KR
6083
65
------
------
---
------
----
------
------
- ---
------
---
------
------
R. p
usio
la
AX61
JQ
8070
93
KJ58
6459
---
------
----
------
------
- ---
------
----
KR67
6909
KR
6083
66
KR67
6787
---
------
----
KR67
6734
KR
6768
37
KR67
6670
R. p
usio
la
AX62
JQ
8070
96
KJ58
6460
---
------
----
------
------
- ---
------
----
KR67
6910
KR
6083
67
------
------
---
------
----
------
------
- ---
------
---
KR67
6671
R. p
usio
la
AX63
JQ
8068
19
JQ80
6798
JQ
8067
77
KR67
6985
---
------
----
KR67
6911
KR
6083
68
------
------
---
------
----
------
------
- ---
------
---
KR67
6672
R. p
usio
la
AY36
KR
6083
15
KR67
7027
---
------
----
------
------
- ---
------
----
------
------
- ---
------
----
------
------
---
------
----
------
------
- ---
------
---
------
------
147
R. p
usio
la
AY38
KJ
5863
78
KJ58
6461
---
------
----
------
------
- ---
------
----
------
------
- ---
------
----
------
------
---
------
----
------
------
- ---
------
---
------
------
R. p
usio
la
AY39
JQ
8070
90
KJ58
6462
---
------
----
------
------
- KR
6769
74
KR67
6912
KR
6083
69
------
------
KR
6770
18
------
------
- ---
------
---
------
------
R. p
usio
la
AZ05
JQ
8070
91
KJ58
6463
---
------
----
------
------
- ---
------
----
KR67
6913
---
------
----
------
------
---
------
----
------
------
- ---
------
---
KR67
6673
R. p
usio
la
BA18
JQ
8070
92
KJ58
6464
---
------
----
------
------
- ---
------
----
KR67
6914
KR
6083
70
------
------
---
------
----
------
------
- ---
------
---
KR67
6674
R. p
usio
la
BA19
JQ
8070
98
KJ58
6465
---
------
----
------
------
- ---
------
----
------
------
- KR
6083
71
------
------
---
------
----
------
------
- ---
------
---
KR67
6675
R. q
ueru
la (s
.s.)
AF16
KJ
5863
80
KJ58
6467
---
------
----
------
------
- ---
------
----
KR67
6916
KR
6083
73
KR67
6789
---
------
----
KR67
6736
KR
6768
39
KR67
6677
R. q
ueru
la (s
.s.)
AG28
KJ
5863
81
KJ58
6468
---
------
----
------
------
- ---
------
----
KR67
6917
KR
6083
74
KR67
6790
---
------
----
KR67
6737
KR
6768
40
KR67
6678
R. q
ueru
la (s
.s.)
AG30
KJ
5863
82
KJ58
6469
---
------
----
------
------
- ---
------
----
KR67
6918
KR
6083
75
KR67
6791
---
------
----
KR67
6738
KR
6768
41
KR67
6679
R. q
ueru
la (s
.s.)
AG32
KJ
5863
84
KJ58
6471
---
------
----
------
------
- ---
------
----
KR67
6920
KR
6083
77
KR67
6793
---
------
----
KR67
6740
KR
6768
43
KR67
6681
R. q
ueru
la (s
.s.)
AG33
KJ
5863
85
KJ58
6472
---
------
----
------
------
- ---
------
----
KR67
6921
KR
6083
78
KR67
6794
---
------
----
KR67
6741
KR
6768
44
KR67
6682
R. q
ueru
la (s
.s.)
AH42
KJ
5863
86
KJ58
6473
---
------
----
------
------
- ---
------
----
KR67
6922
KR
6083
79
KR67
6795
---
------
----
KR67
6742
KR
6768
45
KR67
6683
R. q
ueru
la (s
.s.)
AH51
KJ
5863
87
KJ58
6474
---
------
----
------
------
- ---
------
----
KR67
6923
KR
6083
80
KR67
6796
---
------
----
KR67
6743
KR
6768
46
KR67
6684
R. q
ueru
la (s
.s.)
AJ23
KJ
5863
89
KJ58
6476
---
------
----
------
------
- ---
------
----
KR67
6925
KR
6083
82
KR67
6798
---
------
----
KR67
6745
KR
6768
48
KR67
6686
R. q
ueru
la (s
.s.)
AK05
KJ
5863
90
KJ58
6477
---
------
----
------
------
- ---
------
----
KR67
6926
KR
6083
84
KR67
6800
---
------
----
KR67
6747
KR
6768
50
KR67
6688
R. q
ueru
la (s
.s.)
AL32
KJ
5863
91
KJ58
6478
---
------
----
------
------
- ---
------
----
KR67
6927
KR
6083
85
KR67
6801
---
------
----
KR67
6748
KR
6768
51
KR67
6689
R. q
ueru
la (s
.s.)
AS25
KJ
5863
92
KJ58
6479
---
------
----
------
------
- KR
6769
75
KR67
6928
KR
6083
86
KR67
6802
KR
6770
19
KR67
6749
KR
6768
52
KR67
6690
R. q
ueru
la (s
.s.)
AW30
KR
6083
17
KR67
7029
---
------
----
------
------
- ---
------
----
------
------
- KR
6083
87
KR67
6803
---
------
----
KR67
6750
KR
6768
53
KR67
6691
R. q
ueru
la (s
.s.)
AZ61
JQ
8071
01
KJ58
6480
---
------
----
------
------
- ---
------
----
KR67
6929
KR
6083
88
KR67
6804
---
------
----
KR67
6751
KR
6768
54
KR67
6692
R. q
ueru
la (s
.s.)
BI26
KR
6083
18
------
------
- ---
------
----
------
------
- ---
------
----
------
------
- KR
6083
90
KR67
6806
---
------
----
KR67
6753
KR
6768
56
KR67
6694
R. q
ueru
la (s
.s.)
BI29
KR
6083
19
KR67
7030
---
------
----
------
------
- ---
------
----
------
------
- ---
------
----
KR67
6807
---
------
----
KR67
6754
---
------
---
KR67
6695
R. q
ueru
la (s
.s.)
BI46
KR
6083
20
KR67
7031
---
------
----
------
------
- ---
------
----
------
------
- KR
6083
91
KR67
6808
---
------
----
KR67
6755
KR
6768
57
KR67
6696
R. q
ueru
la (s
.s.)
E56
KR60
8321
KR
6770
32
------
------
- KR
6760
86
------
------
- ---
------
----
KR60
8392
KR
6768
09
------
------
- KR
6767
56
KR67
6858
KR
6766
97
Ravi
nia
stim
ulan
s AD
60
KJ58
6393
KJ
5864
82
------
------
- ---
------
----
------
------
- KR
6769
31
------
------
- ---
------
---
------
------
- ---
------
----
------
------
---
------
---
R. s
timul
ans
AD64
KJ
5863
94
KJ58
6483
---
------
----
------
------
- ---
------
----
KR67
6932
KR
6083
93
KR67
6810
---
------
----
KR67
6757
KR
6768
59
KR67
6698
R. s
timul
ans
AD65
KJ
5863
95
KJ58
6484
---
------
----
------
------
- ---
------
----
KR67
6933
KR
6083
94
------
------
---
------
----
------
------
- ---
------
---
KR67
6699
R. s
timul
ans
AD66
KJ
5863
96
KJ58
6485
---
------
----
------
------
- ---
------
----
KR67
6934
---
------
----
------
------
---
------
----
------
------
- ---
------
---
KR67
6700
R. s
timul
ans
AD67
KJ
5863
97
KJ58
6486
---
------
----
------
------
- KR
6769
76
KR67
6935
---
------
----
------
------
---
------
----
------
------
- ---
------
---
KR67
6701
R. s
timul
ans
AD68
KJ
5863
98
KJ58
6487
---
------
----
------
------
- ---
------
----
KR67
6936
KR
6083
95
------
------
---
------
----
------
------
- ---
------
---
KR67
6702
148
R. s
timul
ans
AE11
KJ
5863
99
KJ58
6488
---
------
----
------
------
- ---
------
----
KR67
6937
KR
6083
96
------
------
---
------
----
------
------
- ---
------
---
KR67
6703
R. s
timul
ans
AE12
KJ
5864
00
KJ58
6489
---
------
----
------
------
- ---
------
----
KR67
6938
KR
6083
97
------
------
---
------
----
------
------
- ---
------
---
KR67
6704
R. s
timul
ans
AR33
JQ
8071
05
KJ58
6490
---
------
----
------
------
- ---
------
----
KR67
6939
---
------
----
------
------
---
------
----
------
------
- ---
------
---
------
------
R. s
timul
ans
AU03
KJ
5864
01
KJ58
6491
---
------
----
------
------
- ---
------
----
------
------
- ---
------
----
------
------
---
------
----
------
------
- ---
------
---
------
------
R. s
timul
ans
AZ10
JQ
8071
09
KJ58
6492
---
------
----
------
------
- ---
------
----
------
------
- ---
------
----
------
------
---
------
----
------
------
- ---
------
---
------
------
R. s
timul
ans
AZ11
JQ
8071
03
KJ58
6493
---
------
----
------
------
- ---
------
----
------
------
- ---
------
----
------
------
---
------
----
------
------
- ---
------
---
------
------
R. s
timul
ans
AZ45
JQ
8071
10
KJ58
6494
---
------
----
------
------
- ---
------
----
------
------
- ---
------
----
------
------
---
------
----
------
------
- ---
------
---
------
------
R. s
timul
ans
AZ46
JQ
8071
11
KJ58
6495
---
------
----
------
------
- ---
------
----
------
------
- ---
------
----
------
------
---
------
----
------
------
- ---
------
---
------
------
R. s
timul
ans
AZ60
JQ
8071
12
KJ58
6496
---
------
----
------
------
- ---
------
----
KR67
6940
KR
6083
98
------
------
---
------
----
------
------
- ---
------
---
------
------
R. s
timul
ans
BA27
JQ
8071
08
KJ58
6497
---
------
----
------
------
- ---
------
----
KR67
6941
KR
6083
99
------
------
---
------
----
------
------
- ---
------
---
------
------
R. s
timul
ans
D45
KJ58
6402
KJ
5864
98
------
------
- ---
------
----
------
------
- ---
------
----
------
------
- ---
------
---
------
------
- ---
------
----
------
------
---
------
---
R. s
timul
ans
E07
GQ22
3320
GQ
2233
20
GQ22
3291
KR
6769
87
------
------
- ---
------
----
KR60
8400
---
------
---
------
------
- ---
------
----
------
------
---
------
---
Ravi
nia
vaga
bund
a BA
15
JQ80
7106
KJ
5864
99
------
------
- ---
------
----
------
------
- KR
6769
42
KR60
8401
KR
6768
11
------
------
- KR
6767
58
KR67
6860
---
------
---
R. v
agab
unda
BA
17
JQ80
7107
KJ
5865
00
------
------
- ---
------
----
KR67
6977
KR
6769
43
KR60
8402
---
------
---
------
------
- ---
------
----
------
------
KR
6767
05
149
Table 4.2. Gene information in multilocus analyses. Name of locus is given with length
(bp), number of sequences (# Seq), number of segregating sites (S), number of
parsimony informative sites (PI), and nucleotide diversity for Ravinia.
Locus bp # Seq. S PI π Mitochondrial DNA
COI 1539 117 69 63 0.06804 COII 700 114 94 83 0.06374 ND4 719 6 101 34 0.06616 ND6 944 7 174 84 0.09938 16S 497 34 24 16 0.01153
Nuclear DNA EF-1a 715 76 8 4 0.01059 G6PD 538 51 84 34 0.02111 ITS2 376 34 12 9 0.02876 PEPCK 401 51 24 16 0.01587 Period 514 47 45 31 0.02235 White 591 74 20 14 0.01428 28s 654 83 2 2 0.00377 Total 8188
150
4.8 FIGURES
Figure 4.1. Sampling map of Ravinia species used for phylogenetic analyses in this study.
Ranges of twelve species are colored according to legend with sampling locations
indicated by stars (see Table S4.1 for specific sampling localities).
152
Figure 4.2. Inferred phylogenetic tree of Ravinia species relationships of North America
using maximum likelihood analysis of 5 mitochondrial genes and 7 nuclear genes.
Support values are given in the following order: Bayesian posterior probabilities
using MrBayes/bootstrap support (%) from ML analysis using RAxML. Shaded
colors represent different Ravinia species. Insert illustrates how the tree topology is
broken up into Sections A and B.
155
Figure 4.3. Wing diagram of a fly indicating the absence or presence of setae on the R1
wing vein (red). Individuals in the Ravinia group lack setae on this wing vein, while
those in the Chaetoravinia group have setae on the R1 wing vein. (Adapted from
McAlpine 1989)
157
Figure 4.4. Species-tree estimation from five mitochondrial (COI, COII, ND4, ND6, 16S) and
seven nuclear loci (EF-1a, G6PD, ITS2, PEPCK, Per, 28S, Wt) using the program
*BEAST. (a) Maximum clade credibility tree generated by TreeAnnotator. Posterior
probabilities are shown above the nodes. (b) Cloudogram of all trees sampled and
visualized by DensiTree.
159
4.9 REFERENCES
Akaike, H. 1973. Information theory and an extension of the maximum likelihood principle.
In: Petrov B.N., Caski, F., editors. Second International Symposium on Information
Theory. Budapest: Akademiai. p. 267-281.
Akaike, H. 1974. A new look at the statistical model identification. IEEE Transactions on
Automatic Control. 19:716-723.
Bouckaert, R.R. 2010. DensiTree: making sense of sets of phylogenetic trees. Bioinformatics
26:1372-1373.
Díaz-Rodríguez, J., Gonçalves, H., Sequeira, F., Sousa-Neves, T., Tejedo, M., Ferrand, N.,
Martínez-Solano, I. 2015. Molecular evidence for cryptic candidate species in Iberian
Pelodytes (Anura, Pelodytidae). Molecular Phylogenetics and Evolution 83:224-241.
DeBry, R.W., Timm, A.E., Dahlem, G.A., Stamper, T. 2010. mtDNA-based identification of
Lucilia cuprina (Wiedemann) and Lucilia sericata (Meigen) (Diptera: Calliphoridae)
in the continental United States. Forensic Science International 202: 102-109.
Drummond, A.J., Ho. S.Y.W., Phillips, M.J., Rambaut, A. 2006. Relaxed phylogenetics and
dating with confidence. PLoS Biology 4:e88.
Drummond, A.J., Suchard, M.A., Xie, D., Rambaut, A. 2012. Bayesian phylogenetics with
BEAUti and BEAST 1.7 Molecular Biology and Evolution 29:1969-1973.
Edgar, R.C. 2004. MUSCLE: multiple sequence alignment with high accuracy and high
throughput. Nucleic Acid Research 32:1792-1797.
Felsenstein, J. 1985. Phylogenies and the comparative method. American Naturalist 125:1-
15.
160
Friedrich, M., Tautz, D. 1997a. An episodic change of rDNA nucleotide substitution rate has
occurred at the time of the emergence of the insect order Diptera. Molecular Biology
and Evolution 14:644-653.
Friedrich, M., Tautz, D. 1997b. Evolution and phylogeny of the Diptera: a molecular
phylogenetic analysis using 28s rDNA sequences. Systematic Biology 46:674-698.
Geyer, C.J. 1991. Markov chain Monte Carlo maximum likelihood. In: Keramides, E.M.,
editor. Computing Science and Statistics: Proceedings of the 23rd Symposium on the
Interface. Fairfax Station, USA, Interface Foundation. p. 156-163.
Gilks, W.R., Roberts, G.O. 1996. Strategies for improving MCMC. In: Gilks, W.R., Richardson,
S., Spiegelhalter, editors. Markov chain Monte Carlo in Practice. London, UK,
Chapman and Hall. p. 89-114.
Giroux, M., Pape, T., Wheeler, T. 2010. Towards a phylogeny of the flesh flies (Diptera:
Sarcophagidae): Morphology and phylogenetic implications of the acrophallus in the
subfamily Sarcophaginae. Zoological Journal of the Linnean Society 158:740-778.
Hebert, P.D.N., Penton, E.H., Burns, J.M., Janzen, D.H., Hallwachs, W. 2004a. Ten species in
one: DNA barcoding reveals cryptic species in the neotropical skipper butterfly
Astraptes fulgerator. Proceedings of the National Academy of Sciences USA
101:14812-14817.
Hebert, P.D.N., Stoeckle, M.Y., Zemlak, T.S., Francis, C.M. 2004b. Identification of birds
through DNA barcodes. PLoS Biology 2:e312.
Heled, J., Drummond, A.J. 2010. Bayesian inference of species trees from multilocus data.
Molecular Biology and Evolution 27:570-580.
161
Kutty, S.N., Bernasconi, M.V., Šifner, F., Meier, R. 2007. Sensitivity analysis, molecular
systematics and natural history evolution of Scathophagidae (Diptera:
Cyclorrhapha: Calyptratae). Cladistics 23:64-83.
Kutty, S.N., Pape, T., Wiegmann, B.M., Meier, R. 2010. Molecular phylogeny of the Calypratae
(Diptera: Cyclorrhapha) with an emphasis on the superfamily Oestroidea and the
position of the Mystacinobiidae and McAlpine’s fly. Systematic Entomology 35:614-
635.
Lanfear, R., Calcott, B., Ho, S.Y.W., Guindon, S. 2012. PartitionFinder: Combined selection of
partitioning schemes and substitution models for phylogenetic analyses. Molecular
Biology and Evolution 29:1695-1701.
Librado, P., Rozas, J. 2009. DnaSP v5: A software for comprehensive analysis of DNA
polymorphism data. Bioinformatics 25:1451-1452.
Lim, G. S., Balke, M., and Meier, R. 2012. Determining species boundaries in a world full of
rarity: singletons, species delimitation methods. Systematic Biology 61:165-169.
Lopes, H.S. 1969. Family Sarcophagidae. In: Papavero, N. A Catalogue of the Diptera of the
Americas South of the United States, Vol. 103. Departamento de Zoologia, Secretaria
da Agricultura, São Paulo. p. 1-88.
Lopes, H.S. 1982. The importance of the mandible and clypeal arch of the first instar larvae
in the classification of the Sarcophagidae (Diptera). Revista Brasileira de Biologia
26:293-326.
Lowenstein, J.H., Amato, G., Kolokotronis, S.-O. 2009. The Real maccoyii: Identifying Tuna
Sushi with DNA Barcodes – Contrasting Characteristic Attributes and Genetic
Distances. PLoS ONE 4: e7866.
162
Maddison, W.P., Maddison, D.R. 2011. ‘Mesquite 2.73: a modular system for evolutionary
analysis.’ Available from [http://mesquiteproject.org].
McAlpine, J.F. (1989) Morphology and Terminology –Adults. In: McAlpine, J.F., Wood, D.M.
editors. Manual of Nearctic Diptera, Vol. 1. Agriculture Canada Research Branch,
Ottawa, CAN. p. 30.
Milne, I., Lindner, D., Bayer, M., Husmeier, D., McGuire, G., Marshall, D.F., Wright, F. 2009.
TOPALi v2: a rich graphical interface for evolutionary analyses of multiple
alignments on HPC clusters and multi-core desktops. Bioinformatics 25:126-127.
Nelson, L.A., Wallman, J.F., Dowton, M. 2007. Using COI barcodes to identify forensically and
medically important blowflies. Medical and Veterinary Entomology 21:44-52.
Pape, T. 1994. The world Blaesoxipha Loew, 1861. (Diptera: Sarcophagidae). Entomologica
Scandinavica Supplement 45:1-247.
Pape, T. 1996. Catalogue of the Sarcophagidae of the World (Insecta: Diptera), Memoirs on
Entomology International, Vol. 8. Associated Publishers, Gainesville, FL.
Pape, T., Dahlem, G.A. 2010. Sarcophagidae (Flesh Flies). In: Brown, B.V., et al., editors.
Manual of Central American Diptera: Volume 2. Ottawa, Ontario, CAN., NRC
Research Press p. 1313-1335.
Rambaut A., Suchard M.A., Xie D., Drummond A.J. 2014. Tracer v1.6, Available
from [http://beast.bio.ed.ac.uk/Tracer].
Roback, S.S. 1954. The evolution and taxonomy of the Sarcophaginae (Diptera,
Sarcophagidae). Illinois Biological Monographs 23:1-181.
Robineau-Desvoidy, J.B. 1863. Histoire naturelle des Dipteres des environs de Paris. Vol. 2.
Paris, France. p. 920.
163
Ronquist, F., Huelsenbeck, J.P. 2003. MrBayes 3: Bayesian phylogenetic inference under
mixed models. Bioinformatics 19:1572-1574.
Saitoh, T., Sugita, N., Someya, S., Iwami, Y., Kobayashi, S., Kamigaichi, H., Higuchi, A., Asai, S.,
Yamamoto, Y., Nishiumi, I. 2015. DNA barcoding reveals 24 distinct lineages as
cryptic bird species candidates in and around the Japanese Archipelago. Molecular
Ecology Resources 15:177-186.
Siebert, S., Backofen, R. 2005. MARNA: multiple alignment and consensus structure
prediction of RNAs based on sequence structure comparisons. Bioinformatics
21:3352-3359.
Song, H., Buhay, J.E., Whiting, M.F., Crandall, K.A. 2008a. Many species in one: DNA
barcoding overestimates the number of species when nuclear mitochondrial
pseudogenes are coamplified. Proceedings of the National Academy of Sciences USA
105:13486-13491.
Song, Z.-K., Wang, X.-Z., Liang, G.Q. 2008b. Species identification of some common
necrophagous flies in Guangdong province, southern China based on the rDNA
internal transcribed spacer 2 (ITS2). Forensic Science International 175: 17-22.
Spillings, B.L., Brooke, B.D., Koekemoer, L.L., Chiphwanya, J., Coetzee, M., Hunt, R.H. 2009. A
new species concealed by Anopheles funestus Giles, a major malaria vector in Africa.
American Journal of Tropical Medicine and Hygiene 81:510-515.
Stamatakis, A. 2014. RAxML Version 8: A tool for Phylogenetic Analysis of Large
Phylogenies. Bioinformatics 30:1312-1313.
164
Stamper, T., Dahlem, G.A., Cookman, C., DeBry, R.W. 2013. Phylogenetic relationships of
flesh flies in the subfamily Sarcophaginae based on three mtDNA fragments
(Diptera: Sarcophagidae). Systematic Entomology 38:35-44.
Sukumaran, J., Holder, M.T. 2010. DendroPy: A Python library for phylogenetic computing.
Bioinformatics 26:1569-1571.
Wiens, J.J., Servedio, M.R. 2000. Species delimitation in systematics: Inferring diagnostic
difference between species. Proceedings of the Royal Society of London B 267:631-
636.
Wong, E.S., Dahlem, G.A., Stamper, T.I., DeBry, R.W. 2015. Discordance between
morphological species identification and mtDNA phylogeny in the flesh fly genus
Ravinia (Diptera: Sarcophagidae). Invertebrate Systematics 29:1-11.
Xiao, J.-H., Wang, N.-X., Li, Y.-W., Murphy, R.W., Wan, D.-G., Niu, L.M., Hu, H.Y., Fu, Y.G.,
Huang, D.W. 2010. Molecular approaches to identify cryptic species and
polymorphic species within a complex community of fig wasps. PLoS ONE 5:e15067.
165
4.10 SUPPLEMENTAL MATERIALS
APPENDIX S4.1. Best Model Selection and Partition Schemes:
Bayesian Inference Analyses (MrBayes)
All Taxa, All Genes: PartitionFinder Input: All genes, all codon position. Greedy algorithm run to determine best-fit model and partitions; resulted in 22 subsets. Scheme lnL = -27455.88508; Scheme AIC = 55823.77016 Subset 1: COI_pos1 = GTR + I + G Subset 2: (COI_pos2, ND4_pos2) = GTR + I + G Subset 3: (COI_pos3, COII_pos3) = GTR + I + G Subset 4: (COII_pos1, ND4_pos1) = GTR + I + G Subset 5: COII_pos2 = HKY + I Subset 6: 28S = GTR + I + G Subset 7: EF1a_pos1 = F81 Subset 8: (EF1a_pos2, PEPCK_pos2, White_pos2) = GTR + I Subset 9: EF1a_pos3 = GTR + G Subset 10: G6PD_pos2 = GTR + I + G Subset 11: G6PD_pos3 = HKY + G Subset 12: (G6PD_pos1) = GTR + I + G Subset 13: ITS2 = HKY + G Subset 14: ND4_pos3 = GTR + G Subset 15: (ND6_pos2, ND6_pos3) = GTR + I + G Subset 16: ND6_pos1 = GTR + I + G Subset 17: PEPCK_pos3 = GTR + G Subset 18: (PEPCK_pos1, White_pos1) = GTR + I + G
Subset 19: (Per_pos1, Per_pos2) = GTR + G Subset 20: Per_pos3 = HKY + G Subset 21: White_pos3 = HKY + G Subset 22: 16S = HKY + I
Maximum Likelihood Analyses (RAxML)
All Taxa, All Genes: PartitionFinder Input: All genes, all codon position. Greedy
algorithm run to determine best-fit model and partitions; resulted in 19 subsets. Scheme lnL = -27466.96967; Scheme AIC = 55863.93934
Subset 1: (COI_1) = GTR + I + G Subset 2: (COI_pos2, COII_pos2, ND4_pos2) = GTR + I + G Subset 3: (COI_pos3, COII_pos3) = GTR + I + G Subset 4: (COII_pos1, ND4_pos1) = GTR + I + G Subset 5: (28S) = GTR + I + G Subset 6: (EF1a_pos1) = GTR + I + G Subset 7: (EF1a_pos2, PEPCK_pos2, White_pos2) = GTR + I + G
166
Subset 8: (EF1a_pos3) = GTR + I + G Subset 9: (G6PD_pos2) = GTR + I + G Subset 10: (G6PD_pos3) = GTR + I + G Subset 11: (G6PD_pos1) = GTR + I + G Subset 12: (ITS2) = GTR + I + G Subset 13: (ND4_pos3) = GTR + I + G Subset 14: (ND6_pos2, ND6_pos3) = GTR + I + G Subset 15: (16S, ND6_pos1) = GTR + I + G Subset 16: (PEPCK_pos3) = GTR + I + G Subset 17: (PEPCK_pos1, White_pos1) = GTR + I + G Subset 18: (Period_pos1, Period_pos2) = GTR + I + G Subset 19: (Period_pos3, White_pos3) = GTR + I + G
*BEAST Analysis
COICOII: (COI_pos1, COI_pos2, COI_pos3, COII_pos1, COII_pos2, COII_pos3) = GTR + I + G, Scheme lnL = -7857.12668; Scheme AIC = 16184.25336
ND4: (ND4_pos1, ND4_pos2, ND4_pos3) = GTR + G = Scheme lnL -825.56456; 1671.12912
ND6: (ND6_pos1, ND6_pos2, ND6_pos3) = GTR + G, Scheme lnL = -3570.56013; Scheme AIC = 7159.12026
EF1a: (EF1a_pos1, EF1a_pos2, EF1a_pos3) = SYM + I, Scheme lnL = -1436.67395; Scheme AIC = 3335.3479
G6PD: (G6PD_pos1, G6PD_pos2, G6PD_pos3) = TN93 (TrN + I + G), Scheme lnL = -2167.08558; Scheme AIC = 4798.17116
PEPCK: (PEPCK_pos1, PEPCK_pos2, PEPCK_pos3) = SYM + G, Scheme lnL = -1289.59556; Scheme AIC = 3041.19112
Period: (Period_pos1, Period_pos2, Period_pos3) = HKY + G, Scheme lnL = -1902.40222; Scheme AIC = 4264.80444
White: (White_pos1, White_pos2, White_pos3) = GTR + I + G, Scheme lnL = -2327.89807; Scheme AIC = 5125.79614
ITS2: = GTR + I + G, Scheme lnL = -838.69565; Scheme AIC = 1697.3913 28S: = GTR + I + G, Scheme lnL -1098.16588; Scheme AIC = 2216.33176 16S: = HKY + I, Scheme lnL = -882.14353; Scheme AIC = 1774.28706
167
SUPPLEMENTAL TABLES
Table S4.1. Specimen sex and locality data. Localities are in the USA and reported as
county and state.
Species I.D. Sex Locality Date Collected
Blaesoxipha arizona (Pape) AW32 Male Grant County, NM 13-Aug-2007
Blaesoxipha cessator (Aldrich) AW27 Male Grant County, NM 13-Aug-2007
Oxysarcodexia cingarus (Aldrich) AP68 Male Schuyler County, NY 09-June-2007
Oxysarcodexia ventricosa (Wulp) E08 Male Hamilton County, OH 10-July-1999
Ravinia acerba (Walker) AU02 Female Kittson County, MN 05-July-2007
Ravinia n.sp. AE79 Male Siskiyou County, CA 07-June-2006
Ravinia n.sp. AE80 Male Siskiyou County, CA 07-June-2006
Ravinia n.sp. AE81 Male Siskiyou County, CA 07-June-2006
Ravinia n.sp. AF03 Male Siskiyou County, CA 07-June-2006
Ravinia n.sp. AG31 Male Jefferson County, CA 08-June-2006
Ravinia n.sp. AG51 Female Jefferson County, OR 14-June-2006
Ravinia n.sp. AH52 Female Jefferson County, OR 14-June-2006
Ravinia n.sp. AJ18 Male Jefferson County, OR 14-June-2006
Ravinia n.sp. AJ19 Male Jefferson County, OR 14-June-2006
Ravinia n.sp. AK04 Male Jefferson County, OR 14-June-2006
Ravinia n.sp. AK25 Female Jefferson County, OR 14-June-2006
Ravinia n.sp. BA16 Male Grant County, NM 13-Aug-2007
Ravinia anxia (s.s.) AT44 Female Kandiyohi County, MN 06-July-2007
R. anxia (s.s.) AT45 Female Kandiyohi County, MN 06-July-2007
R. anxia (s.s.) AT46 Female Kandiyohi County, MN 06-July-2007
R. anxia (s.s.) AT47 Female Kandiyohi County, MN 06-July-2007
R. anxia (s.s.) AT49 Male Kandiyohi County, MN 06-July-2007
R. anxia (s.s.) AT50 Male Kandiyohi County, MN 06-July-2007
R. anxia (s.s.) AW48 Female Grant County, NM 13-Aug-2007
R. anxia (s.s.) AW70 Male Grant County, NM 13-Aug-2007
Ravinia derelicta (Walker) AA10 Female Columbia County, FL 07-May-2006
R. derelicta (Walker) AA13 Female Columbia County, FL 07-May-2006
R. derelicta (Walker) AA25 Female Liberty County, FL 08-May-2006
R. derelicta (Walker) AA26 Female Liberty County, FL 08-May-2006
R. derelicta (Walker) AA38 Female Madison County, FL 08-May-2006
R. derelicta (Walker) AB05 Female Walton County, FL 09-May-2006
R. derelicta (Walker) AB41 Female Washington County, FL 10-May-2006
R. derelicta (Walker) AC75 Female Highland County, FL 12-May-2006
168
R. derelicta (Walker) AC76 Female Highland County, FL 12-May-2006
R. derelicta (Walker) AC81 Male Highland County, FL 12-May-2006
R. derelicta (Walker) AD01 Male Highland County, FL 12-May-2006
R. derelicta (Walker) AD02 Male Highland County, FL 12-May-2006
R. derelicta (Walker) AD06 Male Highland County, FL 12-May-2006
R. derelicta (Walker) AD07 Female Highland County, FL 12-May-2006
R. derelicta (Walker) AD11 Female Highland County, FL 12-May-2006
R. derelicta (Walker) AD40 Male Martin County, FL 13-May-2006
R. derelicta (Walker) AD42 Male Martin County, FL 13-May-2006
R. derelicta (Walker) AE57 Female Glenn County, CA 07-June-2006
R. derelicta (Walker) AN13 Female Jefferson County, TN 22-May-2006
R. derelicta (Walker) AN18 Male Jefferson County, TN 22-May-2006
R. derelicta (Walker) AT48 Female Kandiyohi County, MN 06-July-2007
R. derelicta (Walker) AZ09 Female Hamilton County, OH 12-June-2008
R. derelicta (Walker) AZ62 Female Bland County, VA 19-June-2008
R. derelicta (Walker) BA29 Male Hamilton County, OH 06-July-2008
R. derelicta (Walker) BA30 Male Hamilton County, OH 06-July-2008
R. errabunda (Wulp) AE51 Male Glenn County, CA 07-June-2006
Ravinia floridensis (Aldrich) AA01 Male Duval County, FL 07-May-2006
R. floridensis (Aldrich) AA03 Male Duval County, FL 07-May-2006
R. floridensis (Aldrich) AA04 Female Duval County, FL 07-May-2006
R. floridensis (Aldrich) AA60 Male Sarasota County, FL 11-May-2006
R. floridensis (Aldrich) AB15 Male Washington County, FL 10-May-2006
R. floridensis (Aldrich) AD31 Male Martin County, FL 13-May-2006
R. floridensis (Aldrich) AD41 Female Martin County, FL 12-May-2006
Ravinia lherminieri (R.-D.) AA36 Male Madison County, FL 08-May-2006
R. lherminieri (R.-D.) AN14 Male Jefferson County, TN 22-May-2006
R. lherminieri (R.-D.) AN15 Male Jefferson County, TN 22-May-2006
R. lherminieri (R.-D.) AN16 Male Jefferson County, TN 22-May-2006
R. lherminieri (R.-D.) AN17 Male Jefferson County, TN 22-May-2006
R. lherminieri (R.-D.) AQ58 Female Saratoga County, NY 13-June-2007
R. lherminieri (R.-D.) AZ44 Female Kanawha County, WV 19-June-2008
R. lherminieri (R.-D.) AZ58 Male Kanawha County, WV 19-June-2008
Ravinia planifrons (Aldrich) AE45 Male Klamath County, OR 07-June-2006
R. planifrons (Aldrich) AF05 Male Siskiyou County, CA 07-June-2006
R. planifrons (Aldrich) AF06 Male Siskiyou County, CA 07-June-2006
R. planifrons (Aldrich) AG34 Male Jefferson County, OR 14-June-2006
R. planifrons (Aldrich) AG49 Female Jefferson County, OR 14-June-2006
R. planifrons (Aldrich) AK08 Male Jefferson County, OR 14-June-2006
169
R. planifrons (Aldrich) AK09 Male Jefferson County, OR 14-June-2006
Ravinia pusiola (Wulp) AW64 Female Grant County, NM 13-Aug-2007
R. pusiola (Wulp) AW73 Male Grant County, NM 14-Aug-2007
R. pusiola (Wulp) AX61 Male Grant County, NM 14-Aug-2007
R. pusiola (Wulp) AX62 Male Grant County, NM 14-Aug-2007
R. pusiola (Wulp) AX63 Male Grant County, NM 14-Aug-2007
R. pusiola (Wulp) AY36 Male Grant County, NM 15-Aug-2007
R. pusiola (Wulp) AY38 Female Grant County, NM 15-Aug-2007
R. pusiola (Wulp) AY39 Male Grant County, NM 15-Aug-2007
R. pusiola (Wulp) AZ05 Male Grant County, NM 15-Aug-2007
R. pusiola (Wulp) BA18 Male Grant County, NM 13-Aug-2007
R. pusiola (Wulp) BA19 Male Grant County, NM 13-Aug-2007
R. querula (s.s.) AF16 Male Clackamas County, OR 09-June-2006
R. querula (s.s.) AG28 Male Jefferson County, OR 08-June-2006
R. querula (s.s.) AG30 Male Jefferson County, OR 08-June-2006
R. querula (s.s.) AG32 Male Jefferson County, OR 08-June-2006
R. querula (s.s.) AG33 Male Jefferson County, OR 08-June-2006
R. querula (s.s.) AH42 Male Crook County, OR 08-June-2006
R. querula (s.s.) AH51 Female Crook County, OR 08-June-2006
R. querula (s.s.) AJ23 Female Jefferson County, OR 14-June-2006
R. querula (s.s.) AK05 Male Jefferson County, OR 14-June-2006
R. querula (s.s.) AL32 Male Rabun County, OR 23-May-2006
R. querula (s.s.) AS25 Female Marathon County, WI 02-July-2007
R. querula (s.s.) AW30 Male Grant County, NM 13-Aug-2007
R. querula (s.s.) AZ61 Male Bland County, VA 19-June-2008
R. querula (s.s.) BI26 Male Utah County, UT 05-June-2011
R. querula (s.s.) BI29 Male Utah County, UT 05-June-2011
R. querula (s.s.) BI46 Male Utah County, UT 05-June-2011
R. querula (s.s.) E56 Male Washington County, TN 22-May-2004
Ravinia stimulans (Walker) AD60 Female Volusia County, FL 14-May-2006
R. stimulans (Walker) AD64 Male Volusia County, FL 14-May-2006
R. stimulans (Walker) AD65 Male Volusia County, FL 14-May-2006
R. stimulans (Walker) AD66 Male Volusia County, FL 14-May-2006
R. stimulans (Walker) AD67 Male Volusia County, FL 14-May-2006
R. stimulans (Walker) AD68 Male Volusia County, FL 14-May-2006
R. stimulans (Walker) AE11 Female Newberry County, SC 23-May-2006
R. stimulans (Walker) AE12 Male Newberry County, SC 23-May-2006
R. stimulans (Walker) AR33 Male Saratoga County, NY 13-June-2007
R. stimulans (Walker) AU03 Female Kittson County, MN 05-July-2007
170
R. stimulans (Walker) AZ10 Female Hamilton County, OH 15-June-2008
R. stimulans (Walker) AZ11 Female Hamilton County, OH 15-June-2008
R. stimulans (Walker) AZ45 Female Kanawha County, WV 19-June-2008
R. stimulans (Walker) AZ46 Female Kanawha County, WV 19-June-2008
R. stimulans (Walker) AZ60 Female Bland County, VA 19-June-2008
R. stimulans (Walker) BA27 Male Hamilton County, OH 06-July-2008
R. stimulans (Walker) D45 Female Hocking County, OH 03-July-2003
R. stimulans (Walker) E07 Male Hamilton County, OH 11-July-2003
Ravinia vagabunda (Wulp) BA15 Male Grant County, NM 13-Aug-2007
R. vagabunda (Wulp) BA17 Male Grant County, NM 13-Aug-2007
171
SUPPLEMENTAL FIGURES
Figure S4.1. Species-tree estimation from seven nuclear loci (EF-1a, G6PD, ITS2, PEPCK,
Per, 28S, Wt) using the program *BEAST. Maximum clade credibility tree generated
by TreeAnnotator. Posterior probabilities are shown above the nodes.
173
Chapter Five:
General Conclusions
174
The research that was conducted in this dissertation was motivated by a desire to
explore the evolution of species, the processes involved in speciation, and aspects of
species delimitation using molecular methods. The flesh fly genus Ravinia provided a
model system for exploration in species delimitation using DNA-based techniques. This
study system was chosen because of its broad geographic range, importance in other fields
(i.e., agriculture), presence of longstanding taxonomic disagreement, and ambiguous
species limits based on morphology (Wong et al. 2015). Within Ravinia several species
exhibit limited amounts of diagnostic characters resulting in potential problems when
identifying specimens to species using morphology (Wong et al. 2015). The work presented
within this dissertation is unique in several aspects. This dissertation provides the first
phylogeny for members of the genus Ravinia (Wong et al. 2015; c.f. Chapter 4). In addition,
this research is the first to apply coalescent-based species-tree methods to members of the
flesh fly family Sarcophagidae using a multilocus data set (c.f. Chapter 3). This research is
unique in terms of its integrative molecular approach to species delimitation using
migration, population structure, phylogenetic inference, and coalescent-based analyses
within Sarcophagidae (c.f. Chapter 3).
The first study (Chapter 2) presents a mitochondrial phylogeny of the genus Ravinia,
using the loci: cytochrome oxidase subunits I and II. This is the first phylogenetic study
ever conducted within the genus Ravinia. Previous studies that have included members of
this genus have all been focused on relationships above the species level (Lopes 1982; Pape
1994; Giroux et al. 2010). In Chapter 2, the genetic lineages that are recovered from the
mitochondrial (mtDNA) phylogeny are compared with the morphologically identified
species. The purpose of this comparison was to determine if there was discordance
175
between the mtDNA phylogeny and the morphologically identified species groups. If
discordance was found to exist among species relationships this would provide an impetus
to pursue DNA-based methods of species delimitation within Ravinia. Several paraphyletic
relationships were found to be present among species within this genus. This discordance
between the mtDNA phylogeny and the morphological characters could have been a result
of incomplete lineage sorting, migration and gene flow, cryptic species, or a combination of
the three. However, methods based on a single locus could not differentiate these causes.
As such, a species delimitation study using a multilocus data set would need to be
conducted to determine the cause(s) of this discordance. High support was found for the
groups Chaetoravinia and Ravinia which had previously been separate genera before their
collapse into the currently recognized genus Ravinia (sensu lato). The high nodal support of
the clades along with morphological characters (i.e., R1 wing vein setae) indicates that
these groups are likely naturally occurring and may deserve recognition as separate
genera. However, a formal change in taxonomic status should await further examination of
specimens believed to be part of the previously synonymized genus Adinoravinia.
In the second study (Chapter 3) two species within the genus Ravinia were
delimited: Ravinia anxia (sensu lato) and R. querula (sensu lato). These two species were
selected based on evidence of conflict between the mtDNA phylogeny and the
morphologically identified species-group. In addition, previous taxonomic work had
indicated high morphological variability within R. anxia (sensu lato) and designated this
group as a species complex (Dahlem 1989). A total of 7 molecular markers (2 mtDNA loci
and 5 nuclear loci) were analyzed using a combination of maximum likelihood and
Bayesian inference, migration analyses, population structure analyses, and coalescent-
176
based species-tree analyses. Across all methods the species scenario of R. anxia (sensu lato)
and R. querula (sensu lato), delimited using the current morphological species definitions,
was the least supported. This provides evidence that the current species delimitation is not
accurate for these two species-groups. In general, the species delimitation scenarios
received varying support across methods, with the most highly supported being the 3-
species scenario. Furthermore, the incongruence of species delimitation scenarios between
the methods used could be indicative of important methodological assumptions (i.e.,
migration/gene flow) being violated (Carstens et al. 2013). In systems like Ravinia, with
high levels of genetic diversity and low levels of morphological variation, the accuracy of a
species delimitation study will be dependent on both the data that is available and the
choices that are made by the researchers (Satler et al. 2013). Given the support of the
species delimitation analyses we recognize the three lineages recovered as evolutionary
independent and, as such, distinct species. However, the distinction was only found to be
present within molecular data and additional evidence using other techniques (e.g.,
morphology) would strengthen the status of these species. To prevent the use of nomina
nuda the authors do not revise taxonomic standing; deferring to regulations regarding the
naming of species in the International Code of Zoological Nomenclature (Articles 11-20;
1999). Until future research can produce morphological descriptions the authors
suggested these conventions be used when referring to these genetically delimited species:
Ravinia anxia (sensu stricto) = individuals comprising Clade 1; Ravinia querula (sensu
stricto) = individuals comprising Clade 2; Ravinia n.sp. = individuals comprising Clade 3 (c.f.
Chapter 3).
177
As presented in the third study (Chapter 4), the phylogenetic status of the three
cryptic species (R. anxia (sensu stricto), R. querula (sensu stricto), Ravinia n.sp.), and the
status of the two species Ravinia floridensis and Ravinia lherminieri was investigated using
12 molecular markers (5 mtDNA + 7 nuDNA loci). These markers strongly supported the
distinctiveness of the three cryptic species as monophyletic lineages in the Bayesian
inference, maximum likelihood, and species-tree analyses. This study was unique in that it
was the first to employ a coalescent-based species-tree approach within the flesh fly family
Sarcophagidae using a large data set. In addition, the previously synonymized groups
Ravinia and Chaetoravinia were found to be monophyletic and could merit recognition as
genera. Further studies using members from the Adinoravinia lineage will be needed to
determine if a change in taxonomic status is necessary for these groups.
The work presented by this dissertation is not complete. More research will be
needed, specifically detailed descriptions of the morphological characters, for the cryptic
species to be recognized under the International Coded of Zoological Nomenclature.
However, this research does provide a framework from which to group and compare the
cryptic species in order to identify these morphological characters. This dissertation shows
the impact that gene flow and incomplete lineage sorting (ILS) can have upon the
speciation process. This highlights the importance of taking these events (i.e., gene flow
and ILS) into account when delimiting species and why using an independent, multilocus
data set is critical.
178
REFERENCES
Carstens B.C., Pelletier T.A., Reid N.M., Satler J.D. 2013. How to fail at species delimitation.
Molecular Ecology. 22:4369-4383.
Dahlem GA. “A Revision of the Genus Ravinia (Diptera: Sarcophagidae). Ph.D. Dissertation,
Michigan State University, 1989.
Giroux, M., Pape, T., Wheeler, T. 2010. Towards a phylogeny of the flesh flies (Diptera:
Sarcophagidae): Morphology and phylogenetic implications of the acrophallus in the
subfamily Sarcophaginae. Zoological Journal of the Linnean Society. 158:740-778.
ICZN (International Commission on Zoological Nomenclature) (1999) International code of
zoological nomenclature, 4th edn. London, UK: International Trust for Zoological
Nomenclature.
Lopes, H.S. 1982. The importance of the mandible and clypeal arch of the first instar larvae
in the classification of the Sarcophagidae (Diptera). Revista Brasileira de Biologia.
26:293-326.
Pape, T. 1994. The world Blaesoxipha Loew, 1861. (Diptera: Sarcophagidae). Entomologica
Scandinavica Supplement 45:1-247.
Satler J.D., Carstens B.C., Hedin M. 2013. Multilocus Species Delimitation in a Complex of
Morphologically Conserved Trapdoor Spiders (Mygalomorphae, Antrodiaetidae,
Aliatypus). Systematic Biology. 62:805-823.
Wong, E.S., Dahlem, G.A., Stamper, T.I., DeBry, R.W. 2015. Discordance between
morphological species identification and mtDNA phylogeny in the flesh fly genus
Ravinia (Diptera: Sarcophagidae). Invertebrate Systematics 29:1-11.