Molecular Systematics of the Butterfly Tribe Preponini (Nymphalidae


Material Information

Molecular Systematics of the Butterfly Tribe Preponini (Nymphalidae Charaxinae)
Physical Description:
1 online resource (93 p.)
University of Florida
Place of Publication:
Gainesville, Fla.
Publication Date:

Thesis/Dissertation Information

Master's ( M.S.)
Degree Grantor:
University of Florida
Degree Disciplines:
Entomology and Nematology
Committee Chair:
Willmott, Keith Richard
Committee Members:
Branham, Marc A


Subjects / Keywords:
butterfly -- lepidoptera -- molecular -- preponini -- systematics
Entomology and Nematology -- Dissertations, Academic -- UF
Entomology and Nematology thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation


The first complete species-level phylogeny for the nymphalid butterfly tribe Preponini is presented. In this molecular phylogenetic reconstruction three genes were used, comprising two mitochondrial genes, Cytochrome Oxidase subunits I and II (COI and COII), and the nuclear gene, Elongation Factor ? 1 alpha (EF-1?). Of the three different methodologies that were employed to reconstruct the phylogeny, ML (Maximum Likelihood) provided the highest resolution. This phylogenetic approach allowed me to explore taxonomic limits at different levels. The position of Anaeomorpha splendida Rothschild, 1984 was examined to test tribal limits, with the result that maintaining monophyly of the tribe requires Anaeomorpha?s exclusion. However, the placement of Anaeomorpha within the subfamily Charaxinae is still unresolved, thus a more intensive higher-level phylogeny is required to accurately identify Anaeomorpha?s taxonomic position. A second goal was to investigate generic boundaries, in particular the relationship of Noreppa Rydon, 1971 to Archaeoprepona Fruhstorfer, 1915 and Agrias Doubleday, 1845 to Prepona Boisduval, 1836. In the first case, I synonymize Noreppa with Archaeoprepona syn. nov. since the phylogenetic placement of the former within the latter has high branch support. This revised taxonomy is further supported by morphological data (Marconato 2008), since the same synapomorphies that were formerly used to define Archaeoprepona remain useful after the inclusion of Noreppa. For Agrias and Prepona my results also suggest that these genera should be synonymized. I recommend this merging since Agrias is placed within Prepona and the morphological characters currently used to define these genera show that the same synapomorphies used to identify Prepona apply to Agrias as well. However, it is worth noting that other taxonomic modifications can solve the paraphyly of Prepona as currently conceived. I therefore suggest a detailed morphological study is needed to identify whether there are potential synapomorphies that would support such alternative taxonomic modifications. The use of an additional genetic marker would also be useful to identify and explore the closest relatives to Agrias and provide corroborative evidence before making taxonomic changes. Additionally, I suggest that the description of a new genus for Prepona pheridamas (Cramer, 1777) should be considered, supported by its basal taxonomic placement and differences in morphology in comparison with other Prepona + Agrias (e.g. lack of conspicuous ventral hindwing submarginal eyespots and differences in genitalic characters). To determine whether a new description is the best option, I propose to include an additional marker to increase node support regarding the relationship of Prepona pheridamas to remaining Prepona + Agrias. This study also revealed a number of potential taxonomic issues at lower levels that definitely merit further attention. These include the relationship of Archaeoprepona amphimachus (Fabricius, 1775) to A. meander (Cramer, 1775), Agrias claudina (Godart, 1824) to A. narcissus Staudinger, 1885, broad divergence within Prepona deiphile (Godart, 1824) that might merit recognition as two or more species, and the paraphyly of Prepona pylene Hewitson, 1854. Finally, in addition to examining tribal and generic limits, this study also explored the potential for DNA sequence data to clarify species boundaries within the Prepona laertes (H?bner, 1811) group. Preliminary results do not falsify the hypothesis that Prepona laertes is a polymorphic species with a widespread distribution, since no correspondence between morphology and molecular data was found, and based on the premise proposed by Hebert et al. (2003b) that a 3% genetic divergence is useful for delimiting species. However, the use of such fixed thresholds has been criticized (e.g. Meyer and Paulay 2005), thus suggesting the need for alternative approaches. Nonetheless, when the observed genetic divergences for Prepona laertes were compared to the 9% mean interspecific divergence found for nine preponine species based on the COI sequence data used in the species-level phylogeny, it is evidenced that divergence within P. laertes is still minimal, even in comparison with sister species of Prepona only (4.6 % mean divergence for pair wise comparisons of three species pairs). Nevertheless, there are long branches characterizing several clusters, which might indicate increased genetic distance between such groups, hence the possible presence of cryptic species. However, before drawing any conclusions, it is important to increase sampling effort to rule out artifacts from a lack of sampling. Therefore, I recommend that additional data is required before robust conclusions concerning species identity can be reached. These data should include a broader geographic sampling and a higher number of individuals to help identify stable clusters that can be examined to identify reliable, diagnostic morphological characters.
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility:
by Elena Ortiz.
Thesis (M.S.)--University of Florida, 2011.
Adviser: Willmott, Keith Richard.
Electronic Access:

Record Information

Source Institution:
Rights Management:
Applicable rights reserved.
lcc - LD1780 2011
System ID:

This item is only available as the following downloads:

Full Text




2 2011 Elena Ortiz Acevedo


3 To my family: Rosa, Ana and Juan


4 ACKNOWLEDGMENTS First I would like to thank my family for their constant support. I am deeply thankful to my mother Rosa, for her endless and unconditional love and for supporting me in every step of the way. I want to thank my sister Ana for her good humor and for being the sister anyone would want to have. I am very thankful to you two because you share with me the passion for butterflies and because you encouraged me to follow my dreams. Also, I am very grateful to Juan for his infinite help and for making me part of his life. I know that you learned to love butterflies as much as I learned to love birds. I would lik e to thank my graduate committee, Dr. Keith Willmott and Dr. Marc Branham, for their continuous support through my graduate program. I am very grateful for every comment and suggestion that helped me improve my knowledge and become a better scientist. Ther e are many reasons why I am deeply thankful to Dr. Keith Willmott, his support during my application to the graduate program, guidance through this process, and also because of his great advice and willingness to assist in every step of my graduate studies Finally, I want to thank him for the funding provided to me in order to develop the molecular phase of my project. I want to express my gratitude to Dr. Thomas Emmel for providing a McGuire Center research assistantship that allowed me to come to this amazing place to study, do research and meet incredible people. Also, I want to thank each and every one at the McGuire Center for Lepidoptera and Biodiversity; thanks for making me feel at home. In addition, I want to thank the McGuire center for letting me use its installations and the molecular laboratory. Furthermore, I want to thank every person that collaborated with me and contributed to my research in any way. In particular, I thank Dr. Keith Willmott, Dr.


5 Jackie Miller, Dr. Debbie Mathews Lott, Dr Carmen Pozo (ECOSUR Mexico), Dr. Jorge Llorente (UNAM Mexico), Dr. Dan Janzen (Univeristy of Pennsylvania), Dr. Robert Robbins, Andrew Neild, Werner Zoeller, Christin Brevignn, Mara Fernanda Checa, Geoffrey Gallice, and Cristobal Ros Malaver for giving me access to preponine specimens; Lei Xiao for her help in the molecular laboratory, J. F. LeCrom for his guidance and willingness to teach me about Colombian charaxines; Dr. Paul C. Zalamea and Dr. Pablo Stevenson (Universidad de los Andes Colomb ia) for their logistical and financial support during fieldwork; Universidad de los Andes for giving me access to the Museum ANDES; Sebastin Padrn for the lovely pictures he took of preponine butterflies which are included here and for being a really good friend; Dr. Marta Wolff and Universidad de Antioquia for their logistic support. I want to thank each and every person involved in my fieldwork phase including Juan P. Gomez, Alex, Daniel, Emilse, Ana Jaramillo, Cesar Barrera, Andres Barona, Paul, Pablo, Ari, Judit, Marquinho, Pilar and Conor, and everyone I might be forgetting. I also want to thank Sigma Xis grants in aid of research for a grant awarded to me in 2010, which I used to partially cover fieldwork expenses, the Center for Systematic Entomology for a grant awarded to cover sequencing costs, and the Council of the Linnean Society and the Systematics Association for the Systematics Research Fund award granted to me to cover the remaining molecular costs.


6 TABLE OF CONTENTS page ACKNOWLEDGMENTS .................................................................................................. 4 LIST OF TABLES ............................................................................................................ 8 LIST OF FIGURES .......................................................................................................... 9 ABSTRACT ................................................................................................................... 10 CHAPTER 1 INTRODUCTION .................................................................................................... 13 The Subfamily Charaxinae ...................................................................................... 13 The Tribe Preponini ................................................................................................ 13 Historical Taxonomy ......................................................................................... 14 Current Taxonomy ............................................................................................ 15 Scope of the Project ................................................................................................ 16 2 MOLECULAR PHYLOGENY OF THE TRIBE PREPONINI .................................... 21 Introduction ............................................................................................................. 21 Tribal Limits ...................................................................................................... 21 Preponini Genera ............................................................................................. 22 Materials and Methods ............................................................................................ 26 Sample Acquisition and Origin ......................................................................... 26 Gene Regions .................................................................................................. 27 DNA Extraction and Amplification ..................................................................... 27 DNA Sequencing and Alignment ...................................................................... 28 Phylogenetic Analyses ..................................................................................... 28 Maximum parsimony (MP) ......................................................................... 29 Maximum likelihood (ML) ........................................................................... 29 Bayesian inference (BI) .............................................................................. 29 R esults and Discussion ........................................................................................... 30 3 SPECIES LIMITS WITHIN PREPONA LAERTES .................................................. 43 Introduction ............................................................................................................. 43 Prepona laertes Group ..................................................................................... 43 DNA Barcoding ................................................................................................. 45 Species identification ................................................................................. 48 Species delimitation ................................................................................... 52 Applications of DNA barcoding .................................................................. 54 COI barcoding: difficulties and alternatives ................................................ 56 Using DNA barcoding effectively: the Prepona laertes case. ..................... 61


7 Materials and Methods ............................................................................................ 63 Sample Acquisition and Origin ......................................................................... 63 DNA Extraction and Amplification ..................................................................... 63 DNA Sequencing and Alignment ...................................................................... 63 Cluster Analysis ................................................................................................ 64 Results and Discussion ........................................................................................... 64 4 CONCLUSIONS ..................................................................................................... 78 LIST OF REFERENCES ............................................................................................... 81 BIOGRAPHICAL SKETCH ............................................................................................ 92


8 LIST OF TABLES Table page 2 1 Primers used for DNA amplification .................................................................... 40 2 2 PCR mix protocol used for DNA amplification. ................................................... 41 2 3 PCR cycles for DNA amplification ...................................................................... 42 3 1 Historical classifications of the tribe Preponini. ................................................... 75 3 2 Specimen codes for COI sequence data used in the Prepona laertes study.. .... 76


9 LIST OF FIGURES Figure page 1 1 Representatives of each genus of the tribe Preponini ........................................ 20 2 1 Geographic regions sampled in Colombia .......................................................... 38 2 2 ML inferred tree for the combined data set of COI, COII and EF 1 genes. ....... 39 3 1 The DNA barcode gap. ..................................................................................... 70 3 2 Possible outcomes for the Prepona laertes analysis .......................................... 71 3 3 NJ tree for the Prepona laertes complex. ........................................................... 72 3 4 NJ tree of all available data for Prepona laertes. ................................................ 73


10 Abstract of Thesis Present ed to the Graduate School of the University o f Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science MOLECULAR SYSTEMATICS OF THE BUTTERFLY TRIBE PREPONINI (NYMPHALIDAE: CHARAXINAE) By Elena Ortiz Acevedo August 2011 Chair: Keith Willmott Major: Entomology and Nematology The first complete species level phylogeny for the nymphalid butterfly tribe Preponini is presented. In this molecular phylogenetic reconstruction three genes were used, comprising two m itochondrial genes, Cytochrome Oxidase subunits I and II (COI and COII), and the nuclear gene, E longation F actor 1 alpha (EF 1 ). Of the three different methodologies that were employed to reconstruct the phylogeny, ML (Maximum Likelihood) provided the h ighest resolution. This phylogenetic approach allowed me to explore taxonomic limits at different levels. The position of Anaeomorpha splendida Rothschild, 1984 was examined to test tribal limits, with the result that maintaining monophyly of the tribe r equires Anaeomorphas exclusion. However, the placement of Anaeomorpha within the subfamily Charaxinae is still unresolved, thus a more intensive higher level phylogeny is required to accurately identify Anaeomorphas taxonomic position. A second goal was to investigate generic boundaries, in particular the relationship of Noreppa Rydon, 1971 to Archaeoprepona Fruhstorfer, 1915 and Agrias Doubleday, [1845] to Prepona Boisduval, 1836. In the first case, I synonymize Noreppa with Archaeoprepona syn. nov. since the phylogenetic placement of the former within the latter has high branch support. This


11 revised taxonomy is further supported by morphological data (Marconato 2008), since the same synapomorphies that were formerly used to define Archaeoprepona remain u seful after the inclusion of Noreppa For Agrias and Prepona my results also suggest that these genera should be synonymized. I recommend this merging since Agrias is placed within Prepona and the morphological characters currently used to define these genera show that the same synapomorphies used to identify Prepona apply to Agrias as well. However, it is worth noting that other taxonomic modifications can solve the paraphyly of Prepona as currently conceived I therefore suggest a detailed morphological s tudy is needed to identify whether there are potential synapomorphies that would support such alternative taxonomic modifications. The use of an additional genetic marker would also be useful to identify and explore the closest relatives to Agrias and prov ide corroborative evidence before making taxonomic changes. Additionally, I suggest that the description of a new genus for Prepona pheridamas (Cramer, 1777) should be considered, supported by its basal taxonomic placement and differences in morphology in comparison with other Prepona + Agrias (e.g. lack of conspicuous ventral hindwing submarginal eyespots and differences in genitalic characters). To determine whether a new description is the best option, I propose to include an additional marker to increas e node support regarding the relationship of Prepona pheridamas to remaining Prepona + Agrias This study also revealed a number of potential taxonomic issues at lower levels that definitely merit further attention. These include the relationship of Archa eoprepona amphimachus (Fabricius, 1775) to A. meander (Cramer, 1775), Agrias claudina (Godart, [1824]) to A. narcissus Staudinger, [1885], broad divergence within Prepona deiphile


12 (Godart, [1824]) that might merit recognition as two or more species, and the paraphyly of Prepona pylene Hewitson, [1854]. Finally, in addition to examining tribal and generic limits, this study also explored the potential for DNA sequence data to clarify species boundaries within the Prepona laertes (H bner, [1811]) group. Prel iminary results do not falsify the hypothesis that Prepona laertes is a polymorphic species with a widespread distribution, since no correspondence between morphology and molecular data was found, and based on the premise proposed by Hebert et al. (2003b) that a 3% genetic divergence is useful for delimiting species. However, the use of such fixed thresholds has been criticized (e.g. Meyer and Paulay 2005), thus suggesting the need for alternative approaches. Nonetheless, when the observed genetic divergences for Prepona laertes were compared to the 9% mean interspecific divergence found for nine preponine species based on the COI sequence data used in the species level phylogeny it is evidenced that divergence within P. laertes is still minimal, even in comparison with sister species of Prepona only (4.6 % mean divergence for pair wise comparisons of three species pairs). Nevertheless, there are long branches characterizing seve ral clusters, which might indicate increased genetic distance between such groups, hence the possible presence of cryptic species. However, before drawing any conclusions, it is important to increase sampling effort to rule out artifacts from a lack of sam pling. Therefore, I recommend that additional data is required before robust conclusions concerning species identity can be reached. These data should include a broader geographic sampling and a higher number of individuals to help identify stable clusters that can be examined to identify reliable, diagnostic morphological characters.


13 CHAPTER 1 INTRODUCTION The Subfamily Charaxinae The nymphalid butterfly subfamily Charaxinae contains approximately 400 species within 28 genera (Ackery et al., 1999; Chacn and Montero, 2007) and is distributed throughout the worlds tropical regions. Placed within the satyrine group (Wahlberg et al., 2009) the spectacular wing patterns of many charaxines have made them highly popular among collectors (DeVries, 1987) These butterflies are characterized by a robust body, short and thick proboscis, long palpi (DeVries, 1987) and a rapid flight (Chacn and Montero, 2007) Adults inhabit the canopy and are typically seen only when descending to feed on rotten fruit, dung or carrion (DeVries, 1987) Most species have a brightly colored dorsal surface, whereas the ventral surface is cryptic, resembling dead leaves (Chacn and Montero, 2007; DeVries, 1987) The subfamily includes 5 recognized tribes (Wahlberg et al., 2009), Prothoini, Pallini, Charaxini, Anaeini, and Preponini, with the last two occurring only in the neotropical region. The Tribe Preponi ni The charaxine tribe Preponini, containing 22 currently recognized species, is endemic to the neotropics, where the highest species richness can be found in the Amazon basin (DeVries, 1987) A number of species possess narrow ranges, such as Agrias hewitsonius H. W. Bates, 1860 and Prepona werneri Hering & Hopp, 1925. The brilliant dorsal wing coloration reaches its peak in the genus Agrias the species of which are of high commercial value (Neild 1996) This bright dorsal coloration merits research, because of its high variability (e.g. as shown by Austin (2009) in Agrias claudina and A. amydon Hewitson, [1854]), its putative involvement in mimetic rings


14 between Agrias and Callicore H bner [1819] (Descimon, 1976) and because of its presumed function in sexual signaling. Preponine butterflies have been the focus of numerous studies in many topics, including physiology (Berthier, 2005) ecological studies (Devries and Walla, 2001) and immature stage biology (Furtado, 2001; Muyshondt, 1974; Salazar, 1999) However, despite the fact that thes e spectacular butterflies have been the subject of various taxonomic treatments (Johnson and Descimon, 1988, 1989; Llorente Bousquets et al., 1992) there still remain major questions to be addressed in terms of bot h generic and species level systematics. Recent studies that have included only a handful of representatives have found problems with defining the tribe, the genera Prepona and Archaeoprepona, and in delimiting some of the most common species (Freitas and Brown, 2004; Marconato, 2008; Wahlberg et al., 2009) A large part of the problem is the relative morphological homogeneity among species, providing few characters of taxonomic value. Hence, analysis from the molec ular perspective seems promising for the group. However, before beginning any systematic study it is important to review the current and historical taxonomy of the group to provide a broad knowledge about its classification and thus develop appropriate met hodologies. Histor ical Taxonomy The first species descriptions of preponines were published in the mid18th century. All species described at that time were placed within Papilio and are now included within the genus Archaeoprepona; A. demophon (Linnaeus, 1758), A. amphimachus (Fabricius, 1775), A. meander (Cramer, 1775) and A. licomedes (Cramer, 1777). P reponine species are distributed within five genera (Figure 11), with the first member of Archaeoprepona described by Linnaeus in 1758. The genus Prepona was


15 introduced in 1836 by Boisduval, followed by Doubleday's introduction of Agrias in 1845, a genus that only differs from the former in wing pattern and minor variations in the antennae and genitalia according to Fruhstorfer (1916) Later, in 1894, Anaeomorpha was establis hed by Rothschild, positioning it as a genus that stands mid way between Anaea H bn. and Prepona Boisd., followed by Fruhstorfer's Archaeoprepona in 1915. Finally, Noreppa was introduced by Rydon in 1971 based on the eyes possessing hairs, and on diffe rences in both genitalia and midand hind tibia and tarsi (Neild, 1996) Since the introduction of the various preponine genera the taxonomy of Preponini has been stable. However, a detailed study of the groups taxonomy is lacking. Several studies in butterfly systematics have included preponine represen tatives, which serve as stepping stones for further research and for resolving taxonomic conflict within the group. Current Taxonomy Existing phylogenetic studies of Preponini have been done only as part of higher level phylogenetic studies of butterflies. Brower (2000) based on partial sequences of the nuclear gene wingless assessed t he phylogenetic relationships among the tribes of the family Nymphalidae. Even though he included Preponini, just one representative of the tribe was taken into account, Prepona sp T he most parsimonious cladogram yield ed a good resolution for the subfamil y Charaxinae, placing the tribe Preponini sister to all other charaxines In a later work carried out by Freitas and Brown (2004) a phylogenetic hypothesis of the family Nymphalidae wa s proposed based on morphological characters of each of the life stages. However, this generic level phylogeny only had one representative of Preponini, Archaeoprepona chalciope


16 ( H bner, [1823]) Marconato (2008) presented a phylogeny of the subfamily Charaxinae based on morphological c h aracters, including the tribe Preponini, which was represented by 13 species She concluded that maintaining the monophyly of the tribe requires that Anaeomorpha splendida be excluded from the group, which agrees with the suggestion made by Pea and Wahlberg (2008) and Wahlberg et al. (2009) that this species is more closely related to the tribe Anaeini. Wahlberg et al. (2009) proposed a phylogenetic hypothesis of 400 genera of butterflies, which included one species of four out of the five recognized genera for the tribe, and provi ded more support for Marconatos findings. Scope of the Project Further progress in preponine taxonomy requires a comprehensive approach, in terms of both taxa and characters. According to Mayr (1963) the use of independent data sets, such as morphological and molecular data, has been considered important for reconstructing patterns of relatedness at both higher and lower taxonomic levels. Baker et al. (1998) showed that using both molecular and morphological data sets for a wide range of taxonomic groups, including plants, insects, reptiles and primates, had a positive effect on the combined analysis as showed by partitioned Bremer support. On the other hand, independent data sets of genomic origin (i.e. different genes) can also be used to obtain more robust results as shown by Rokas et al. (2003) when combining 106 gene regions of eight yeast species. Fi nally, Miller et al. (1997) and Wahlberg et al. (2005) showed the usefulness of combining independent data sets in phylogenetic analyses of Lepidoptera in particular. To date, in butterflies, morphology has been key in classification (Ackery et al., 1999) at different taxonomic levels. The most intensively studied characters include


17 wing venation, color pattern and genitalia. Wing venation has been extensively used in the classification of insects at the order and family level (Combes and Daniel, 2003) and is also widely employed in the Lepidoptera (DeVr ies, 1987) at family (Triplehorn and Johnson, 2005) genus (LeCrom et al., 2004) and species levels (Elias et al., 2007) However in many cases, venation characters should be reinforced with other morphological characters when delimiting groups (Albrecht and Kaila, 1997) For butterfly species identification, color patterns are virtually alway s useful, and the use of color plates for identification is almost universal (e.g. Bravo et al., 2009; DeVries, 1987; Neild, 1996) Color patterns may also be converted into standard identification keys (e.g. LeCrom et al., 2004) However as noted by Bickford et al. (2007) in some cases morphology might not provide reliable characters, for example due to strong selection for intra or interspecific sexual signaling, and/or selection against character divergence. In Preponini, wing venation might not be a helpful character in distinguishing genera since it is uniform across several genera (Fruhstorfer (1916) recalling Schatz), such as Prepona and Agrias Instead, Fruhstorfer (1916) suggested that the color of the hindwing anal margin androconial tufts could be used to identify the two main groups in the tribe (further supported by differences in their genitalia), namely the genera Archaeoprepona and Prepona + Agrias However, he acknowledged the potential limitation of such coloration charact ers, since at least some South American species show geographic variation in the color of their androconial tufts. The limitation of this character was also discussed by Marconato (2008) Another morphological character which might not reliably distinguish sibling species in Preponini is color pattern, because it apparently shows high variation within


18 sp ecies, as shown by Austin (2009) for Agrias (noted earlier by Fruhstorfer (1916) ), and as noted by Neild (1996) for Prepona. At the generic level, color pattern is the main character used to distinguish Prepona and Agrias ( Fruhstorfer (1916) despite the fact that genitalia show no consistent differences between the genera. Prepona praeneste Hewitson, 1859 exemplifies a species intermediate in wing pattern between the two genera Agrias and Prepona, and indeed Furtado (2008) demonstrated hybridization between the two genera. The close relationship between Agrias and Prepona was discussed by Marconato (2008) who suggested that Agrias represented an unusually colorful group of Prepona. At higher levels, alt hough both morphological (as evidenced above) and molecular characters suggest problems with preponine taxonomy (Marconato, 2008; Wahlberg et al., 2009) existing studies have not been sufficiently taxonomically com prehensive to resolve these. Therefore, my goal was to use molecular sequence data to attempt to resolve major taxonomic issues in the Preponini group through the first species level phylogenetic approach. Furthermore, the molecular phylogenetic results c an be compared with previous phylogenies, both molecular and morphological, to identify areas of congruence or conflict, and to gain a better understanding of character evolution. The resulting phylogeny will be an initial step to a possible future revision of genera within the tribe, in particular Agrias and Prepona. To explore the potential of molecular characters to help in resolving species complexes, a cryptic species group will also be the focus of my research to improve our knowledge of the potential diversity and taxonomy of the group.


19 Specifically my aims are: 1. Define the limits of the tribe Preponini, in particular examining the taxonomic position of Anaeomorpha splendida. 2. Test the monophyly of all preponine genera. 3. Assess species limits within th e Prepona laertes group.


20 Figure 11. Representatives of each genus of the tribe Preponini. From upper left towards lower right: Agrias Anaeomorpha, Archaeoprepona, Prepona, and Noreppa. Coloration pattern on the left side corresponds to the dorsal view the right side corresponds to the ventral view. Photographs by Pablo Sebasti n Padrn.


21 CHAPTER 2 MOLECULAR PHYLOGENY OF THE TRIBE PREPONI NI Introduction Preponine species have been included in several recent systematic studies, but most of these have concentrated on higher level taxonomic issues. However, Marconato (2008) focused her cladistic analysis on the subfamily Charaxinae, thus providing a finer resolution of the phylogenetic relationships within the group. She highlighted several key points in the topology that should be further studied: i) the relationship of Anaeomorpha splendida to other Preponini, with implications for tribal limits, and ii) the relationships of Noreppa and Agrias to other genera, with implications for generic classification. Tribal Limits The representation of preponine butterflies in phylogenetic studies has increased ov er time. Freitas and Brown (2004) proposed a phylogeny of the family Nymphalidae, including only one species of Preponini, Archaeoprepona chalciope. In both of their phylogenetic reconstructions, this species appear ed as a sister species to the charaxine tribe Anaeini Later, Pea et al. (2006) included Archaeoprepona demophon as the only representative of Preponini in their molecular study of Satyrinae. In the topology presented for the combined dataset of three genes Archaeoprepona demophon appeared as sister to Hypna clytemnestra (Cramer, 1777), within the tribe Anaeini, with the latter taxon appearing as a paraphyletic group. Browers (2000) phylogenetic study of the Nymphalidae based on t he nuclear gene wingless placed the only included representative of the tribe Preponini, Prepona sp within a polytomy with members of the tribes Charaxini and Anaeini. After successive approximations weighting the


22 resolution improved, leaving Prepona sp as sister species to a group formed by the tribes Charaxini and Anaeini. Later, Pe a and colleagues (2008) found no support for the monophyly of the tribe in a study that incl uded the genera Agrias, Prepona, Archaeoprepona and Anaeomorpha. In the Bayesian Inference topology the monotypic genus Anaeomorpha appeared as sister to the only representative of the African tribe Pallini, with Anaeomorpha + Pallini being the sister grou p to the tribe Anaeini. Consistent with Pe a et al.s (2008) results were those of Wahlberg et al. (2009) who reconstructed the phylogeny for approximately 74% of recognized Nymphalidae genera based on 10 genes and 235 morphological characters, also omitting the genus Noreppa. The maximum likelihood analysis showed Anaeomorpha as sister to Anaeini, whereas Agrias, Prepona, and Archaeoprepona formed a monophyletic group. Lastly, Marconato (2008) developed the only phylogenetic study cont aining representatives of all five Preponini genera. In her cladistic analysis the most parsimonious tree placed Anaeomorpha within a polytomy with Anaeini representatives. After successive approximations weighting, Anaeomorpha was again placed outside Preponini, and as sister species to the monophyletic Anaeini. Preponini Genera As described above for Preponini tribal limits, generic limits within the tribe have also been debated, and even though morphology has been used to distinguish Preponini genera (Fi gure 1 1) since their descriptions, some relationships remain problematic. The identification of Prepona and Archaeoprepona is straightforward and based on four principal characters (Johnson and Descimon, 1988) These characters w ere also analyzed by Marconato (2008) but while they distinguished Archaeoprepona


23 from Prepona they failed to differen tiate in some instances among Noreppa, Anaeomorpha and Agrias (see below). The first character discussed by Johnson and Descimon (1988) for identification of preponine genera is the modified male scent scales or androconia. In Archaeoprepona and Noreppa these patches of long, hair like scales are diffuse and the hairs flexible, located along the anal margin of the dorsal hindwing. In contrast, Prepona and Agrias possess well defined and rigid brush like setae located in the same position. In addition to the morphology, the coloration of the androconial scales can be used to identify each genus, being dark in Archaeoprepona and yellow in Prepona and Agrias (Marconato, 2008) Marconato (2008) suggested the presence of androconial scales of this form a s a synapomorphic character for the tribe Preponini except for the genus Anaeomorpha which lacks this character. ii) Small and undifferentiated ventral hindwing su bmarginal eyespots in the genera Archaeoprepona and Noreppa in comparison with the large eyespots present in Prepona. This character was divided into two characters by Marconato (2008) who suggested that having eyespots in Rs M1 and Cu1Cu2 equal in size and shape to other marginal hindwing eyespots was plesiomorphic, thus the conspicuous eyespots in Pre pona can be considered synapomorphic. However, Prepona pheridamas lacks conspicuous eyespots, sharing the ancestral characteristic of Archaeoprepona. Marconato (2008) also addressed the size of the marginal ventral hindwing eyespots in the genera Archaeoprepona and Agrias suggesting that the plesiomorphic character state was small eyespots in Archaeoprepona (with the exception of Archaeoprepona chalciope which lacks eyespots ) in contrast to larger eyespots as in the genus Agrias


24 Even though she did not address the state of this character in Noreppa, I would s uggest that this genus also possesses the plesiomorphic character state. O ther diagnostic character s are found in the male genitalia (iii), which is characterized by a flat gnathos lacking spines and with the genitalic capsule being more robust in Archaeoprepona whereas in Prepona the gnathos is rodlike with the presence of spines and the genitalia are more slender overall. In Marconatos (2008) morphological study she scored variation within the gnathos as two independent characters. First, the degree of sclerotization was found to differ betw een Archaeoprepona and Prepona, being heavily sclerotized in the former (also including Noreppa ) and moderately sclerotized in the lat ter (including Agrias ). L ack of sclerotization was found to be the plesiomorphic character state The second character addressed was the morphology of the gnathos. She suggested four character states in which the lack of texture (either spines or striations) was considered plesiomorphic. For this character the Archaeoprepona + Noreppa clade possessed striations (including a r evers al in Archaeoprepona amphimachus ) whereas the Prepona + Agrias clade possessed a high number of spines. Interestingly, the gnathos of Prepona p h eridamas possessed several spines, and this species appeared as basal to Prepona + Agrias (Marconato, 2008) The last character proposed by Johnson and Descimon (1988) refers to the female sterigma, which is Y shaped in Prepona and circular in Archaeoprepona. Marconato (2008) als o addressed this character but proposed three character states. The plesiomorphic character state was found to be the Y shaped fused lamella antevaginalis found in the Agrias amydon clade only. Archaeoprepona possessed the


25 apomorphic character state, whi ch is V shaped. Finally, Noreppa possessed the autopomorphy of having the fused lamella antevaginal is U shaped. For this character the genus Prepona was not scored Noreppa, even though easily differentiated from the above genera ( Archaeoprepona and Prepona ) due to the presence of orange spots near the dorsal hindwing margin (Neild, 1996) appeared within Archaeoprepona in Marconatos (2008) study. As explained above, for some characters this genus is not differentiated from Archaeoprepona, such as androconial shape and color, size and location of marginal hindwing eyespots, and gnathos sclerotization and texture (male genitalia), thus suggesting that Noreppa should perhaps be synonymized with Archaeoprepona. Overall Marconatos (2008) morphological study, which included 6 members of Prepona and Agrias found that Agrias amydon and Agrias claudina appeared within Prepona. Furtado (2008) provided additional evidence for the close relationship between Agrias and Prepona by successfully pairing a female Agrias claudina godmani Fruhstorfer, 1895 with a male of Prepona omphale ( H bner, [1819]) (currently Prepona laertes (Hbner, [1811]) according to Lamas, 2004) In summary, studies to date have shown that more comprehensive work is needed to define the tribal and generic limits in Preponini. A more thorough study of the tribes phylogenetic relationships should help to choose taxonomic solutions depending on the phylogeny of included taxa. Hall and Harveys (2001) cladistic analysis of the genera Juditha Hemming, 1964, Lemonias H bner, [1807] Thisbe H bner, [1819] and Uraneis H. W. Bates, 1868 (Lepidoptera: Riodinidae: Nymphidiini) is a good example of the several possible relationships which may be found among taxa within genera, and of the


26 potential solutions to disentangle the results. Since monophyletic clusters are required in a phylogenetic classification they overcame paraphyly and polyphyly by using synonymy in some cases as well as describing a new genus. In the present study, I use a comprehensive molecular phylogenetic approach to suggest the most robust and appropriate taxonomy for the tribe Preponini. Materials and Methods Sample Acquisition and Origin To maximize the number of species and specimens for this study I carried out fieldwork in Colombia, and contacted collaborators in other neotropical countries. Fieldwork involved sampling in two seasons, from May Ju ly 2010 and two weeks in the month of December 2010. Colombia promi sed to be a useful country for sampling because it comprises different habitats and biogeographic regions with a correspondingly high number of representatives of the tribe; 19 species out of 22 according to Lam as (2004) I sampled several regions, including the two interandean valleys (Cauca and Magdalena) P acific coast Choc Llanos/P lains, an d Amazonian rainforest (Figure 21 ). To collect preponine butterflies I used baited traps that were hung from the canopy and left there for a minimum of six days. They were freshly baited with rotten fish or fermenting fruit every 48 h, and were checked every day for the whole collecting period. Butterflies were stored inside glassine envelopes for transport, with two legs removed and placed in 96% EtOH, and finally spread in the laboratory. In addition, I received recently collected material, both papered and pinned, from collaborators in Mexico, Honduras, Germany, United Kingdom, Venezuela, and Ecuador. I also obtained some tissue samples from the collection at the McGuire Center for Lepidoptera and Biodiversity, Gainesville, FL, which included specimens from Costa


27 Rica and Brazil, as well as from visits to the Smithsonian Institution National Museum, Washington D.C., and Museo de Zoolog a Alfonso L. Herrera at Universidad Nacional Aut noma de M xico, M xico D.F. Gene Regions For the molecular phylogeny of Preponini I sequenced two mitochondrial genes, COI and COII, and one nuclear gene, EF locus analyses provide increased support for phylogenetic hypotheses (Brito and Edwards, 2009; Liu et al., 2008; Wahlberg and Wheat, 2008) and these genes have proven of value in phylogenetic studies of butterflies (Hundsdoerfer et al., 2009; Kandul et al., 2004; Silva Brando et al., 2005) DNA Extraction and Amplification I extracted DNA from two legs either i) removed from freshly killed specimens and stored in 96% EtOH or ii) recently removed from dried specimens stored in glassine envelopes or spread. I extracted DNA from butterfly tissue using th e Qiagen DNEasy Extraction Kit and following their protocol. For old tissue samples the extraction protocol was modified following I udica et al. (2001) M odifications included prolonged digestion with proteinase K for approximately 36 48 h, Ph test of the DNA sample before adsorption on the QIAGEN column (pH higher than 7.0 decreases adsorption) and final elution in 50ul of elution buffer. I used standard Polymerase Chain Reaction (PCR) to amplify the target genes. The COI barcode region of approximately 650bp in length was amplified following Hebert et al. s (2004a) PCR protocol with modifications and using the primers LEPF1 (f) and LEPR1 (r) also as described by these authors In cases where the previous primers were not successful the primers LCO (f) and HC O (r) were used (Folmer et al., 1994)


28 The COII mitochondrial gene of approximately 1000bp in length was amplified following Hillis et al. s (1996) protocol, and using the primers George (f) and Eva (r) (Monteiro and Pierce 2001) Finally, the nuclear gene EF 1 of approximately 1000bp in length was amplified following Hillis et al.s (1996) protocol and the primers EF44 (f) and EFrcM4 (r) (Monteiro and Pierce, 2001) (primer names and sequences available in table 21). The PCR mix used was the same for all g ene regions, which is described in table 2 2. PCR protocols for each gene are detailed in table 2 3. DNA Sequencing and Alignment Custom DNA Sequencing from both strands of each gene was carried out by University of Floridas Interdisciplinary Center for B iotechnology Research Sanger Sequencing Group. Geneious v5.3 (Drummond et al., 2010) was used to manually edit both strands of each gene, with the resulting sequences aligned us ing ClustalW (Larkin et al., 2007) and a consensus sequence produced. The final sequence obtained for COI was 618bp in length, 897bp for COII, and 955bp for EF 1 Subsequently, I constructed a matrix combining these three loci for a total length of 2470bp. Phylogenetic Analyses The complete data set for the three genes was analyzed under the Maximum Parsimony (MP) optimality criterion using PAUP* 4.0 (Swofford, 2003) .The data were also analyzed by Maximum Likelihood (ML) using Randomized Axelerated Maximum Likelihood BlackBox (R AxML BlackBox) (Stamatakis, 2006; Stam atakis et al., 2008) and under the Bayesian Inference (BI) criterion using Mr. Bayes 3.1 (Huelsenbeck et al., 2001; Ronquist and Huelsenbeck, 2003) both implemented in the Cipres Portal


29 ( / ). To edit the resulting trees I used the program FigTree v1.3.1 (Rambaut, 20062009) Maximum parsimony (MP) Phylogenetic tree estimation was carried out by a heuristic search in which characters were equall y weighted and gaps were treated as missing data, using the tree bisection reconnection (TBR) branch swapping algorithm and a simple stepwise addition. A strict consensus tree was estimated when multiple equally parsimonious trees were obtained, followed by a bootstrap analysis (Felsenstein, 1985) as a measure of branch support and re liability of the resulting tree (Hall, 2008) This analysis involved a full heuristi c search including 1000 replicates and retaining groups with frequency 50%. Maximum likelihood (ML) The ML analysis involved a partitioned analysis in which the sequence evolution model for each gene was selected using the program jModelTest (Posada and Crandall, 1998) The model GTR + G was chosen for all data sets. T he phylogenetic tree estimate was followed by rapid bootstrap analysis (Stamatakis et al., 2008) consisting of 1000 replicates. Bayesian inference (BI) The model of sequence evolution chosen by jModelTest (Posada and Crandall, 1998) was GTR + G for the whole data set of 2470bp as well as for each individual gene. BI analyses involved bot h a partitioned and a nonpartitioned analysis. Four Markov Chain Monte Carlo simulations of 200,000 generations sampling every 100 generations with a burn in of 25% of generations were executed. A majority rule consensus tree with branch posterior probabi lities was constructed for both analyses.


30 Results and Discussion A total of 43 individuals from all 22 species recognized in the tribe by Lamas (2004) were included in this study. From this group a total of 23 individuals possessed the complete gene dataset, only two genes were successfully amplified for nine individuals, and the remaining eight individuals had DNA data for just one gene. The outgroup species represented the charax ine tribe Anaeini, and the subfamilies Brassolinae and Satyrinae for a total of 11 species from which only two had an incomplete gene dataset (lack ing sequence for EF 1 ). The ML tree constructed for the combined data of the three genes is shown in Figure 2 2. The topology produced by the ML analysis showed the best resolution for the group. However, MP (including bootstrap analysis) and BI analyses were also carried out, and even though the topologies were not well resolved they highly supported several nodes which are denoted by and + respectively in Figure 22. The ML analysis was designed to test two major issues of Preponini systematics. The first issue is the monophyly of the tribe, particularly the placement of Anaeomorpha splendida (top right Fi gure 11). Rothschild, in 1984, suggested that Anaeomorpha splendida was related to the tribe Anaeini, and other authors, based on phylogenetic studies, have come to similar conclusions (Wahlberg et al., 2009) or sh own a relationship to the tribe Pallini (Pea and Wahlberg, 2008) Recently, Marconatos (2008) study left Anaeomorpha as a member of Preponini, despite the fact that Anaeomorpha splendida does not possess androconial tufts which are considered a synapomorphy for Preponini according to her analysis. Her results showed Anaeomorpha to be placed in a polytomy with representatives of Pallini, Anaeini,


31 Preponini and Charaxini, though after successive approximations weighting it was placed as sister to Anaeini. The results from the ML analysis supp ort the suggestion that Anaeomorpha splendida is misplaced within the tribe Preponini, and thus the monophyly of the latter requires Anaeomorphas exclusion. The clade formed by both Ecuadorian specimens shows high bootstrap and posterior probability values and is placed as sister clade to Anaeini + Preponini; however, this position does not have significant branch support. Contrary to what was proposed by Wahlberg et al. (2009) and Marconato (2008) these results show that Anaeomorpha, instead of being more closely related to the tribe Anaeini, may actually fall outside the current boundaries of the tribe Anaeini, a result consistent with P ea and Wahlbergs (2008) suggestion that Anaeomorpha is more closely related to the tribe Pallini. Therefore, I suggest that Anaeomopha splendida should be excluded from Preponini and placed in an incertae sedis until its placement within the charaxines is further analyzed. Future work in this aspect should include representatives from other tribes, particularly from Pallini. In addition, in order to increase the node support, I would suggest broadening the gene pool, by adding another nuclear gene, such as CAD IDH or RpS5 which are useful markers in this kind of study (Wahlberg, pers. comm.). The second objective of this analysis was to test the monophyly of the Preponini genera. In particular, two main issues involving preponine genera were suggested as requiring further research by Marconato (2008) : i) the placement of Noreppa chromus (Gu rin Mneville [1844]) with respect to the genus Archaeoprepona and ii) relationship of Agrias with the genus Prepona. Escalante et al. (2010) further


32 investigated these issues but focused only on Mexican preponines. Therefore the present study, besides representing a broader geographic dimension, includes all preponine species, which increases the accuracy and adds robustness to the results. The first issue regarding Preponini generic classification involves the potential placement of Noreppa within Archaeoprepona In Figure 2 2 it is clear that Noreppa chromus is placed within the genus Archaeoprepona, and is sister to A. licomedes Its well supported position, and the relative ly low differentiation among other Archaeoprepona species suggest that Noreppa should be considered as a representative of Archaeoprepona. Additional support for this merging can be found in Marconatos (2008) morphology based phylogeny. According to her results, the synapomorphies that characterize Archaeoprepona also occur in Noreppa Thus, even though there are other taxonomic modifications that could solve this conflict, for instance maintaining Noreppa and describing additional genera for other monophyletic groups, the genetic evidence coupled with the morphological data suggest that placing Noreppa w ithin Archaeoprepona is the best solution. Therefore, I here synonymize the genera Noreppa syn. nov. and Archaeoprepona. Also of interest within the Archaeoprepona cluster is that individuals of A. amphimachus and A. meander form a mixed group, and even though the MP, ML and BI node support are not significant, the ML analysis suggests that these species should be further studied. Attempting to complete the full set of genes for the individuals analyzed here should be a priority and could prove helpful in improving node support. The potential merging of both species was suggested by Escalante et al. (2010) who used COI barcode data to assess the phylogenetic relationships among Mexican


33 preponines. According to their results and due to the low genetic distance between both species they propose their merging. However, they only included one individual from Archaeoprepona meander, thus yielding a result that could potentially be an artifact of the lack of samples. Further research on this topic should be attempted in the future, including more individuals from different subs pecies, in order to disentangle the real evolutionary relationships among these taxa. I also highlight the need for additional genetic data for the species Archaeoprepona demophoon (H bner, [1814]). My results show that Archaeoprepona demophoon is sister to remaining species of Archaeoprepona clade, but the low node support and fact that only one individual was included (which lacked EF 1 data), suggest that this placement may require confirmation My results contrast to Marconatoss (2008) cladistic analysis in which Archaeoprepona chalciope was sister to remaining Archaeoprepona, followed by A. demophon and then A. demophoon as sister species to the group formed by A. licomedes + ( A. chromus + A. amphimachus ). Therefore, broadening the phylogenetic breadth in terms of individuals from Archaeoprepona demophoon and DNA data might prove to be useful in corr oborating my current results. Moreover, a remaining issue to be considered regarding the former genus Noreppa involves the status of the Colombian taxon priene (Hewitson, 1859), currently treated as a subspecies of A. chromus but formerly regarded as a di stinct species. I had only a single individual available for study, and I was not successful in obtaining DNA data. Therefore, future analyses should also aim to address this issue.


34 The second issue involving preponine generic classification regards the r elationship between the genera Agrias and Prepona. Several important facts are apparent in Figure 22. Firstly, Agrias representatives form a well defined clade within Prepona, although the nearest relatives of Agrias are still unclear since significant MP bootstrap support and BI posterior probability are lacking. In contrast, the internal nodes of the Agrias clade have significant support. This cluster is composed by two major subdivisions, one of which is the clade formed by Agrias narcissus, A. claudina, and A. aedon Hewitson, 1848. This clade suggests that Agrias claudina is not monophyletic, with both individuals of Agrias claudina lugens Staudinger, 1886 split, with one of them showing a higher affinity to A. narcissus. Increasing the phylogenetic scope to test for experimental error would be useful to confirm this result, since including just one individual from a particular species in a species level study could yield inaccurate results (Funk, 1999) In particular, more individuals of Agrias narcissus with a full set of genes shou ld be included, since the only individual of A. narcissus included lacked EF 1 data, and one individual of Agrias claudina lugens lacked COI data. If Agrias claudina continues to appear as a paraphyletic species, this may reflect ancient polymorphism (Avise and Ball, 1990) or introgression resulting from interspecific hybridization (Sota and Vogler, 2001) The successful pairing of a male of Prepona with a female Agrias by Furtado (2008) suggests that interspecific mating between A. narcissus and A. claudina is certainly possible. The second clade within Agrias is formed by Agrias hewitsonius and A. amydon, a group for which all the methodologies yielded significant node support. Even t hough the ML analysis proved to be useful in clarifying the relationships within the genus


35 Agrias several species were represented by just one individual (e.g. A. aedon) and some individuals lacked DNA data. Improving both of these aspects would yield mor e solid results in future studies. Moving towards deeper nodes in the topology, the Agrias clade is placed as sister to the group formed by Prepona praeneste and P. deiphile neoterpe Honrath, 1884. Prepona deiphile neoterpe is the sister species to P. pra eneste, a clade supported by high bootstrap and posterior probability values. However, Prepona deiphile appears as polyphyletic since individuals from a different subspecies, P. deiphile ibarra Beutelspacher, 1982, group with P. pylene thus forming a group sister to Agrias + ( P. praeneste + P. deiphile neoterpe). These results suggest that P. deiphile might constitute two different species. However, the node support is very low for all three methodologies, suggesting the use of additional markers to increas e support for these findings. Similarly, Prepona pylene appears as paraphyletic, a result that despite having high node support merits further study by adding more specimens as well as more markers. The results shown here suggest that both subspecies might be independent entities, potentially different species. The remaining species of the genus Prepona, namely P. werneri, P. laertes, P. dexamenus Hopffer, 1874, and P. pheridamas appear as well established groups within the Prepona + Agrias cluster. Even though these four species have high bootstrap node support, the inclusion of more individuals of P. laertes might increase the posterior probability values to 0.95 as in the case of P. werneri, P. dexamenus and P. pheridamas The star in Figure 22 denotes the presence of splitting that has very small branch length (3 x 10-6) thus appearing as a polytomy.


36 The ML topology clearly shows that Prepona as currently conceived is paraphyletic, and even though there are several possible taxonomic solutions, one of these seems by far the most appropriate since it requires fewer taxonomic modifications and is more reasonable in view of the current classification of other charaxine genera. I recommend that Agrias should be synonymized with Prepona since, despite the fact that Agrias species form a well defined cluster, they are placed well within the Prepona clade. Another possible taxonomic solution would be expanding the currently conceived Agrias to include the taxa deiphile, praeneste, pylene and werneri a cla de supported by high ML bootstrap values, in addition to describing new genera for the highly supported taxa dexamenus and pheridamas and keeping Prepona as a monotypic genus ( Prepona laertes as its single representative). Nevertheless, when combining the available morphological and genetic information, the synonymy of Agrias seems the most taxonomically reasonable solution. This is due to the fact that the morphological synapomorphies that define Prepona, with the exception of Prepona pheridamas (discusse d below), can also be used to define Agrias according to Marconatos (2008) cladistic analysis. Desc ription of a new genus for Prepona pheridamas should also be considered, since its morphology (e.g. lack of conspicuous eye spots on the ventral hindwing and structure of the gnathos in the male genitalia) differs from Prepona and its placement is basal to the Prepona + Agrias clade. These results also support Marconatos findings based on morphological data. The gnathos morphology (i.e. possessing several spines) of Prepona pheridamas is unique (Marconato (2008) and thus would provide an autopomorphy for a new genus. However, I would also suggest a more detailed


37 morphological study addressing additional possible characters that could be used to better define a new genus. In conclusion, before a more detailed morphological study aiming to identify synapomorphic characters for the potential putative new genera (e.g. for the taxon dexamenus ) is available, I would suggest that species from Agrias and Prepona should be considered as congeneric. In addition, and as suggested above for other cases, I think that adding an extra nuclear marker would be useful to provide corroborating evidence prior to taking such a taxonomic step.


38 Figure 21. Geographic regions sampled in Colombia. Pacific Coast Choc Region (blue), Interandean Valley, Cauca (pink) and Magdalena (green), Llanos/Plains (yellow), and Amazon Rainforest (purple).


39 Figure 22. ML inferred tree for the combined data set of COI, COII and EF 1 genes. Numbers above branches correspond to rapid bootstrap values 50% from the ML analysis. Asterisks (*) correspond to nodes with > 50% bootstrap support from the MP analysis and sum symbols (+) represen 0.95 posterior probability in the BI analysis. Star symbol denotes a splitting of branches.


40 Table 2 1. Primers used for DNA amplification. Both primers are shown in the 5 3 direction. Primers Gene Forward Reverse COI (Hebert et al., 2004a) (Folmer et al., 1994) LEP -F1 ATTCAACCAATCATAAAGATAT LCO TAAACTTCAGGGTGACCAAAA AATCA LEP -R1 TAAACTTCTGGATGTCCAAAAA HCO GGTCAACAAATCATAAAGATATTGG COII (Monteiro and Pierce, 2001) George ATACCTCGACGTTATTCAGA Eva GAGACCATTACTTGCTTTCAGTCATCT EF -1 (Monteiro and Pierce, 2001) EF44 GCYGARCGYGARCGTGGTATYAC EFrcM4 ACAGCVACKGTYTGYCTCATRTC


41 Table 2 2. PCR mix protocol used for DNA amplificat ion. The same protocol was used for the three genes. Asterisks indicate that in some cases the amount of DNA was increased to 4ul in which case the amount of ddH2O was decreased to 10.6ul. Substance Volume (ul) Buffer 10x 2 dNTPs (2.5uM) 0.8 MgCl 2 (50m M) 0.8 Primer Forward 0.4 Primer Reverse 0.4 ddH 2 O 12.6* Taq Platinum 1unit/ul DNA 2* Total 20


42 Table 2 3. PCR cycles for DNA amplification. Asterisks indicate that each cycle has three different steps: DNA denaturation, primer annealing, and DNA elongation, each with its specific time and temperature. Gene Number of cycles Time (seconds) Temperature (Celsius) COI 1 60 94.0 5* 60 94.0 30 45.0 40 72.0 35* 30 94.0 40 51.0 40 72.0 1 300 72.0 COII and EF 1 35* 60 95.0 60 58.0 60 72.0 1 300 72.0


43 CHAPTER 3 SPECIES LIMITS WITHI N PREPONA LAERTES Introduction Prepona laertes Group Preponini species level taxonomy has traditionally been a highly debated topic. The most recent Neotropical butterfly checklist (Lamas, 2004) recognized twenty two species in the tribe, within five genera: Prepona containing seven species, Archaeoprepona with eight, Agrias with five, and Noreppa and Anaeomorpha both monotypic. Neild (1996) stated that there are approximately n ine species within Archaeoprepona, and two within Noreppa, but remained uncertain about the number of species in Prepona and Agrias His treatment of Prepona was based on his own observations in the field and in collections, but his study was limited to Venezuela and environs. He suggested that the real number of species in Agrias was likely to lie between four to nine, as previously stated by Barselou (1983) and D'Abrera (1987) respectively. Table 3 1 summarizes various taxonomic proposals for each of the Preponini genera over time. It is evident that these pr oposals are remarkably variable, in particular for the genus Prepona, in a large part due to the taxonomy of the Prepona laertes group. The variation in color pattern within this species group has resulted in an extraordinary number of taxonomic descriptio ns since the beginning of the twentieth century, when LeMoult (1932 in VaneWright, 1974) proposed approximately 70 new names. Neild (1996) recalled personal communication from Lamas that species within the Prepona laertes complex should be considered one polymorphic species, including P. omphale and P. laertes that were f ormerly divided by Fruhstorfer (1916) This


44 treatment was reflected in the latest checklist by Lamas (2004) who suggested four subspecies for the complex which comprises the former species known as P repona omphale, P. joiceyi, P. pseudojoiceyi, P. phili p poni, P. rothschildi, P. lilianae, and P. pseudomphale (all described by LeMoult, 1932 with the exception of P. omphale), amongst others. Neild (1996) however, considered Prepona phi li p poni from the Orinoco, Amazon, Guianas, and southern Brazil and P. pseudomphale, from Venezuela, Guianas and the Amazon, to be species distinct from the widespread P. laertes Polymorphism is commonly seen in other butterfly groups such as Nymphalinae and Heliconiinae, and is often the result of mimicry (Joron et al., 1999; Linares, 1996) Studies to date have been based on qualitative consideration of ecology, geographic distribution and morphology. Neild (1996) identified morphological characters from wing pattern and androconial tufts for Prepona laertes as well as for P. philipponi and P. pseudomphale. According to Neild, Prepona laertes is characterized by yellow to yellow orange androconial tufts compared to the brownish red coloration of P. pseudomphale and brownishorange in P. philipponi Th is character, however, is somewhat variable and its assessment in some specimens is very subjective. Neild also gave wing characters that he suggested were useful in identification of these three putative species. The shape of the black line located in the median section on the ventral hindwing is a helpful character, which is straight or slightly curved in Prepona philipponi in contrast to curved and irregular in P. pseudomphale and curved and irregular to a lesser extent in P. laertes. In addition, Prepona philipponi hindwing coloration shows a sharp contrast between the creamy ochre portions of the wing, a contrast that is not so strong in the remaining two species. To separate males of P.


45 pseudomphale and P. laertes it is useful to examine ventral hindw ing cells M2 Rs in which there is a creamy colored space between the postdiscal black line and the ochre portion of the wing that is absent in P. laertes Regarding females, Neild (pers. comm.) confirmed that they are very difficult to identify. Despite th is, he suggested useful characters for the Prepona laertes complex females from Venezuela and the Guianas, such as the presence of blue coloration in spaces four, five and six in the dorsal forewing of P. philipponi females, a coloration that is absent in females of P. pseudomphale. Since such morphological characters vary from region to region and are somewhat subjective, an independent approach, such as molecular study, could provide an important new set of data to help disentangle the relationships and s pecies limits within the Prepona laertes group. DNA Barcoding DNA barcoding is a novel method that use s a short easily amplified gene region to universally identify species. Identification can be achieved by either assigning an unknown sequence to a pa rticular species by comparing the sequence against a matrix of knowns or by hypothesizing the discovery of cryptic species. P roponents of the technique point out several problems of traditional taxonom ic methods that can be reduced by using molecular methods For example, morphology might be of limited value in identifying species when phenotypic plasticity is present or, on the contrary, when phenotypic characters are invariant (e.g. cryptic species) (fully described by Hebert et al., 2003b) In addition, the current use of morphological keys, which is common in traditional taxonom ic revisi ons requires significant knowledge of the organisms and sometimes can be limiting in terms of gender or life stage (Hebert et al., 2003b)


46 DNA barcoding, which has set itself the target of identify ing global biodiversity has been the center of ongoing debate. The short segment of mtDNA ( Cytochrome Oxidase subunit I (COXI or COI) in the case of animals) from individuals of interes t is analyzed and compared to a mtDNA matrix based on distance methods, yielding infor mation about sequence variation within the sequence pool (Hebert et al., 2003b) F or the DNA barcode approach to work, either all species must be reciprocally monophyletic (Wiens, 2002) or intraspecific variation must be lower than interspecific divergence (Meyer and Paulay, 2005) t he so called barcoding ga p (Figure 31). For species diagnosis, Hebert et al. (2004b) proposed the 10x rule in which the average sequence divergence between sequence clusters should be greater than ten times divergence within clusters. Other, somewhat arbitrary thresholds may also be chosen (e.g. Meyer and Paulay, 2005) such as mean interspecific divergence, but such thresholds can lead to errors (Meier et al., 2 008) This issue is critical according to Wiemers and Fiedler (2007) who did not find a barcoding gap in Agrodiaetus H bner, 1822 (Lepidoptera: Lycaenidae), a group in which interspecific divergences were as low as 0%, and in which many closely related species showed low interspecific divergence yielding an overlap of 18% in COI sequence divergence when intra and interspecific sequences were compared. DNA barcoding might assist in two main areas of systematic biology (Sites and Marshall, 2003; W iens, 2007) : i ) discover and describe new species, and ii ) assess their phylogenetic relationships. However, the utility of DNA barcoding in terms of these two major goals remains unresolved, with co nflicting opinions generating a continuing debate. Concerns about sequence variation and how to interpret them to delimit


47 species, the distance methods employed, substitution rates of mtDNA, exhaustive sampling including broader areas and sister species, and the need for a priori hypothes e s are examples of the nonagonist side of the coin (e.g. Brower, 2006; Meyer and Paulay, 2005; Moritz and Cicero, 2004; Prendini, 2005; Will and Rubinoff, 2004) that need to be taken into account when using the technique. Regardless of t h e se criticism s, DNA barcodes help in the quest of finding at least some species boundaries and their identification (Hebert and Gregory, 2005) even if they might not provide sufficient evidence for species description (but see Brower, 2010) The majority of researchers suggest that a comprehensive, integrated data analysis (fully described by Will et al., 2005) is the best approach, requiring independent data sets (Baker et al., 1998; Mayr, 1963; Miller et al., 1997; Rokas et al., 2003; Wahlberg et al., 2005) their combination (Ferguson, 2002; Rubinoff and Holland, 2005) and the development of innovative concepts and applications such as the taxonomic circle (DeSalle et al., 2005) in which t he interactions between data sets is assessed to reveal new species. However, in some situations nonmolecular data sets such as morphology might not be helpful in systematic research, as pointed out by Bickford and colleagues (2007) Where morphology is suspected to be intraspecifically variable, the DNA barcoding technique is a powerful tool (Elias et al., 2007; Magnacca and Brown, 2010) Nevertheless, attempting to obtain a more comprehensive matrix s hould not be left to one side. There is ever increasing research on DNA barcoding, and its achievements in a wide group of organisms have been well documented. T he purpose of this review is to synthesize the information availabl e on Lepidoptera, and some other groups of


48 arthropods I aim to address the two main goals of the method by examining specific cases in arthropods, butterflies and moths. In these sections I will also focus on important critiques of the technique and how they have been assessed in studies of Lepidoptera. Some important applications of the technique will be reviewed. Subsequently, I will address further insights into how to use the technique efficiently in the case of Prepona laertes Finally, I will discus s several difficulties associated with and alternatives to this molecular tool. Species identification One of the main goals of DNA barcoding is to assist in species identification. As explained in the previous section, the mechanism by which COI sequences from unidentified organisms can be linked to particular taxa requires a data matrix containing DNA data from a pool of individuals from the same and other species to which the unknown sequence can be compared and matched using distance methods that comp are sequence variation (Hebert et al., 2003b) Importantly, as comprehensive a matrix as possible is required to obtain ac curate results (Ekrem et al., 2007; Little and Stevenson, 2007) Brower (2006) debated whether DNA barcoding could permit species identification. He state d th at instead of concluding that a seque nce from an unknown taxon belongs to X species, one can only say that the sequence is more similar to another than to the remaining sequences in the matrix, because one is relying on % of sequence variation. The fact that DNA barcoding relies on distance methods to generate an identification was examined by Little and Stevenson (2007) They compared different algorithms when assessing gymnosperm identification, and their results showed contrasting results depending on the algorithm used. They showed that hierarchical c lustering methods such as Neighbor Joining (NJ), Parsimony and


49 Maximum Likelihood performed the worst ; these rely on comparing known and unknown sequences in a database, and finding a topology after their alignment. On the other hand, similarity methods addressed such as BLAST performed better; these methods yield a similarity and probability score after comparing the unknown sequence against an unaligned matrix using a pai r wise partial alignment. Regardless of these findings, clustering methods such as NJ have been and continue to be widely used with barcode sequence data to identify unknown sequences at least in Lepidoptera (e.g. Hebert et al., 2004a) Despite the continuing debate regarding the benefits and disadvantages that DNA barcoding yields, it has been shown to be a fair technique to approach species identification at different lev els. When introducing the barcoding concept Hebert et al. (2003a) evaluated identification of unknowns at phyla, order (insect), and species level. This was achieved by gathering a data set of 100 sequences (one per species), 100 sequences (one per family), and 200 sequences (closely related species) respectively. The resulting NJ topology based on the analysis of amino acids coded by COI for seven phyla showed mixed results; there was good resolution for some groups but not for others (e.g. Arthropoda vs. Mollusca respectively). At t he ordinal level there was better resolution, with all orders monophyletic except Coleoptera. Also, they assess ed COI sequence variation and tested assignations of unknowns to three taxonomic categories, with a high rate of success; 100% at the ordinal and species level and 96.4% at phylum level. Within Lepidoptera there has been extensive work done o n DNA barcoding partly because this group supposedly poses a rigorous test for the technique because of its


50 high diversity and the low sequence divergence amongst families (Hebert et al., 2003a) Lepidoptera contains almost 160, 000 described species and is amongst the five most speciose insect orders (Gullan, 2004) S tudies have evaluate d biodiversity under the barcode perspective, and some assess specific criticisms of the method. Janzen et al. (2005) found the DNA barcoding technique to be successful in identifying members of the complex and diverse Lepidoptera fauna of the Area de Conservacin Guanacaste, Costa Rica Hajibabaei et al. (2006) provided similar results for the same Costa Rican region, in which approximately 98% of 521 species belonging to the families Hesperiidae, Sphingidae, and Saturnii dae were successfully identified. Later, Elias et al. (2007) tested the techniques performance in a diverse group of butterflies with a solid taxonomy based on morphology. They showed that t he barcoding technique might not be as effective as previously stated, at least in complex tropical faunas, with only 44 of 57 species identified based on molecular data. The number successfully identified decreased when intraand interspecific variation increased due to inclusion of geographically distant samples. Since the technique relies o n clustering sequences according to their resemblance to each other, monophyly of species would seem to be important. However, Lukhtanov et al. (2009) suggested that monophyly wa s not required for species identification in a sample of Palearctic butterflies Th ey found that the separation of monophyletic groups from paraphyletic groups was highly supported by bootstrap analysis and that such groups form nonoverlapp ing clusters Second ly if a cluster is formed by monophyletic and paraphyletic parts (i.e. two species), one can still identify a diagnostic barcode sequence for the paraphyletic part since it shares


51 synapomorphies with the monophyletic part and at the s ame time lacks the synapomorphies present in the monophyletic remainder. The monophyletic portion would have a barcode sequence characterized by the synapomorphies shared with the paraphyletic part plus its own synapomorphies. Despite the fact that sever al authors have suggest ed that broader geographic sampling is needed in order to increase interspecific and intraspecific variation (e.g. Meyer and Paulay, 2005) Lukhtanov et al. (2009) found that the success of identifying species was not drastically reduced when the geographic scale was broadened. They studied approximately 70% of the central Asian butterfly fauna comprising 353 species, 370 populations, and 880 COI sequences, which they compared using the Kimura 2parameter yielding a NJ tree further analyzed by bootstrap testing. They explicitly address ed the influence on identification success of including geographically separated populations by analyz ing the genetic distance between pairs of closely related species when: i) including one population, and ii) including additional populations. Their results show that increasing the geographic breadth did not reduce the success in identifying unknowns because increasing intraspecific variation does not fill the interspecies hiatus. Other evaluations (e.g. Prendini, 2005) of the barcoding approach (e.g. Barrett and Hebert, 2005; Hebert et al., 2004b) have noted that barcoding studies typically exclude close relatives (usually allopatric) and therefore miscalculat e error rates (Meyer and Paulay, 2005) Meyer and Paulay (2005) tackle d the importance of geographic variat ion and the distribution of species, suggesting that both could pose particular problems to the method. They call ed attention to the fact that comprehensive sampling needs to be


52 achieved for DNA barcoding to be successful, in terms of increasing: i) inters pecific variation through an extensive geographic survey, and ii) intraspecific variation by means of increasing the individuals analyzed. In their study of cowries (Gastropoda, Cypraeidae) they show proof of the importance of a broad geographic sampling. W hen assessing a regionally restricted fauna the barcoding gap proved to be higher (lower error rates), but as an artifact of the exclusion of allopatric sister taxa. However, they acknowledge that each model organism might have different geographic scal es in which error reduction occurs, thus the extent of a comprehensive sampling will depend on the organism assessed. Species delimitation Another main goal of DNA barcoding is to provide an important tool to help in studies of species delimitation as an initial approach (Hebert et al., 2004a) Sometimes research focused on identifying sequences of unknown individuals can evolve into studies involving cryptic species complexes or delimiting s pecies. In such cases the results obtained from a routine identification process might reveal a cluster of unknowns within the matrix where divergence among them is lower than with respect to remaining clusters As explained in the first section the im plementation of thresholds is a useful approach when using the technique to analyze sequence data for species delimitation. H ere, contrary to the case of species identification, in species delimitation I believe that it is crucial that putative species are monophyletic Several studies regarding this topic and involving Lepidoptera have been published, such as Hebert et al. (2003b) This study, in addition to provid ing more support for the techniques success in Lepidoptera by being able to differentiate closely related species of butterf lies and moths, provided an estimate of the mean divergence


53 value for pai rs of closely related species in the order, namely 6.6%. Later, Burns et al. (2008) demonstrated the utility of DNA barcoding when assessing cryptic species complexes in Costa Rican Hesperiidae by providing evidence for four cryptic species within Perichar es philetes (Gmelin, [1790]) However, contrasting results regarding the discrimination of closely related species of Lepidoptera have also been published (e.g. Kaila and Stahls, 2006) As well as showing s uccess for the technique in Lepidoptera, Hebert and colleagues (2004a) provided a controversial example of how DNA barcoding can help in assessing cryptic species complexes in a s tudy of the skipper butterfly Astraptes fulgerator (Walch, 1775) Here they combined their genetic results with ecology and natural history traits to provide a strong case for the existence of cryptic species suggesting that the former A fulgerator compr ised a group of 10 cryptic species However, several critiques of this work have been published since, particularly Brower (2006) who presented a rigorous evaluation of the paper by reanalyzing the data by means of other methods. He called attention to three main elements of Hebert et al.s work, which in his opinion, needed to be reevaluated. First he mentioned the fact that clade support was not assessed, thus we cannot be certain about the stability of the groups. Second, he analyze d DNA sequence data variation using other methods such as population aggregation analysis and cladistic haplotype analysis to corroborate the topology provided by Hebert et al. (2004a) He f ound some contradictions between these results and those originally presented. Third, he noted that no out group was used, which would be needed to make an evolutionary calibration of the data. Finally,


54 he suggested that a broader sampling effor t would be needed to incorporate all possible mtDNA haplotypes corresponding to Astraptes fulgerator before drawing conclusions It is difficult to establish a fixed amount of genetic divergence at which different clusters should be considered different species. In my opinion, the amount of divergence needed to treat clusters as different species depends on several factors such as the groups inherent characteristics for genetic divergence. For example, Meyer and Paulay (2005) noted that marine gastropods tended not contain allopatric sist er taxa, whereas the opposite was usually true for terrestrial gastropods, where dispersal among regions was much more limited. Hebert et al. (2003a; 2003b) found low sequence divergence in the COI region in the phy lum Cnidaria, and suggested that this might be partially explained by the fact that their mitochondria possess an excision repair mechanism that is absent in other groups. Since increasing geographic sampling tends to increase divergence within species clu sters, the threshold divergence for species recognition should depend on the extent of geographic sampling. In general, broader geographic sampling would require lower divergence between clusters to qualify for species recognition. On the contrary, narrower geographic sampling would demand more divergent clusters. Additionally, where possible, information other than DNA sequence divergence should be used in making taxonomic decisions. For example, Furtado (2008) suggested that two Preponini genera, Agrias and Prepona, should be combined since their species were ab le to hybridize Applications of DNA barcoding DNA barcoding can be used to address other biological questions in addition to species discovery and identification. Hajibabaei et al. (2007) assessed the pot ential use


55 of DNA barcode regions and related techniques such as DNA microarrays for monitoring biodiversity in mammals. They noted the importance of scale in determining which method of study is most appropriate. Large scale studies would benefit from COI barcoding because this sequencebased approach can discover unknown haplotypes by comparing the unknown sequence data to previously identified organisms. With DNA microarrays, however, with a broad scale of study the appearance of new haplotypes might com plicate the analysis because this technique requires previous knowledge of target sequences to be analyzed. Darling and Blum (2007) reviewed the use of DNA barcoding for identifying and monitoring invasive species. They highlight ed the value of barcoding when addressing unknown samples but also emphasize d the potential difficulties with more complex scenarios such as surveying overall diversity in a specific area. In those complex cases they suggested that using other methodologies (e.g. DNA fingerprinting) would be more appropriate. Smith et al. (2006) tested the use of DNA barcoding to address the cryptic species complex of parasitoid flies reared from lepidopteran caterpillars in Costa Rica and advocated the use of this method in developing strategies of biological control in related groups. Cywinska et al. (2006) and Rivera and Currie (2009) successfully applied DNA barcoding in medical studies. S ome studies have implications for biodiversity conservation in natural areas where biodiversity is impacted, either by anthropogenic factors or by nonnative fauna displacing native species Other c onservation approaches using DNA barcoding are common, although the goals are not always clear (Moritz 1994). Petit et al. (1998) used genetic markers to assess allelic richness and thus determine how best to protect a population of the argan tr ee ( Argania spinosa (L.) Skeel ) in Morocco. A number of other


56 studies on the usefulness of genetic markers i n biodiversity conservation and monitoring have been published for different organisms (e.g. DeSalle, 2006; Fral, 2002; Gaudeul et al., 2000; Rubinoff, 2006; Schwartz et al., 2007) Within Lepidoptera, Go mpert et al. (2006) used genetic data (m t DNA) in conservation and management of the Karner blue butterfly, Lycaeides melissa (W. H. Edwards, 1873) but highlight ed the need to complement the results obtained with di fferent markers for the reliable diagnos is of conservation units. COI barcoding: difficulties and alternatives DNA barcoding has been the su bject of much research, yield ing insights into the dynamism of species, their relationships with each other and the ir boundaries. However, research has also created awareness of the limitations and difficulties involved with the technique, not only because of theoretical and empirical issues discussed above, but also because of external factors. Examples of some of the latter difficulties include: i) heteroplasmy, ii) pseudogenes, and iii) Wolbachia Hertig, 1936. i) Heteroplasmy can be defined as an organism having more than one mitochondrial DNA haplotype (Magnacca and Brown, 2010) ; thus species showing heteroplasmy might be equivocally identified or remain unknown. Magnacca and Brown (2010) tested the effect of heteroplasmy on the accuracy of DNA barcoding for species identification of bees in the genus Hylaeus Fa bricius, 1793. The y found that approximately 43% of a priori morphological species sampled were heteroplasmic. DNA barcoding identified 100% of homoplasmic species but only 71% of heteroplasmic species However, they also noted that success in identifying h eteroplasmic species could be improved (in this case from 71% to 86%) by treating polymorphisms as character states.


57 This study also found the presence of pseudogenes (ii). Pseudogenes (numts) are DNA sequences that resemble functional genes but are not expressed (Li et al., 1981) Hence, the presence of numts in the organisms being tested either for identification or for species delimitation can yield equivocal results due to co amplification of the pseudogene with the mitochondrial DNA functional gene. Numts have been studied in several organisms including insec ts. Bensasson et al. (2000) showed that migration of mitochondrial DNA and its assimilation in the nucleus is frequent i n grasshoppers, a phenomenon further examined by Bensasson et al. (2001) and reviewed by Timmis et al. (2004) Song et al. (2008) evaluated the effect of numts on DNA barcoding success in two different lineages of arthropods. They found that numts we re co amplified with the COI gene. In grasshopper s the ir analysis found 17 clusters based on a threshold of 3% sequence divergence, which means that 13 more species were detected in addition to the four currently recognized. As a conclusion, they suggest future research is needed to detect the possible presence of numts in order to avoid overestimating the number of species or compromis ing their co rrect identification. Sorenson and Quinn (1998) reviewed different techniques used to detect numts. First, one could clone the PCR products followed by sequencing. Also, the use of internal primers to the origi nal sequence can be used to differentiate between both sets of sequences. The use of restriction endonucleases to digest one of the copies is another approach that can be used. In addition, one can use mRNA of the gene product to obtain a cDNA copy of the expressed mtDNA. Finally, the use of the southern blot technique using the DNA isolated against a purified mtDNA probe is a more straightforward way to detect numts.


58 Another feature that might jeopardize studies based on mitochondrial DNA is the presence of the maternally inherited endosymbiont Wolbachia ( iii). This bacterial endosymbiont is found in many arthropod lineages and involved in reproductive alterations such as c ytoplasmic incompati bility and parthenogenesis (Hoy 2003) Its evolution and phylogene tic relations have been analyzed (e.g. Werren et al., 1995b; Zhou et al., 1998) as well as several of its biological aspects such as its ability to transfer horizontally by means of parasitism (Dedeine et al., 2005; Heath et al., 1999) its capability of transferring its genetic material to the hosts DNA (Kondo et al., 2002) its potential advantage to its host against viral infections (Hedges et al., 2008) and its possible benefit against insect disease transmission (Brownstein et al., 2003) Some studies have evaluated the involvement of Wolbachia in insects, for instance its effect on Drosophil a Fall n, 1823 (Shoemaker et al., 2004) and Solenopsis Westwood, 1840 (Shoemaker et al., 2003) mtD NA evolution. Other studies have assessed its taxonomic distribution (e.g. Werren et al., 1995a; Werren and Windsor, 2000) while others have studied its role in barcoding approaches. Hurst and Jiggins (2005) addressed W olbachia effects on phylogeography, phylogeny and barcoding studies, and conclude d that a Wolbachia infection might lead to a mistaken inference of evolutionary history Similarly, Narita et al. (2006) examined the prevalence of Wolbachia in the butterfly Eurema hecabe (Linnaeus, 1764) populations in Japan. Their results suggest that there are two sibling species within the recognized Eurema hecabe. These two forms were s upported by nuclear and biological data. On the contrary, the mtDNA data did not support the former results due to the infection of the populations with Wolbachia. A later study by Whitworth et al. (2007) in the genus Protocalliphora


59 Hough, 1899 (Diptera: Calliphoridae) found evidence for the possible introgression of Wolbachia genetic material, which introduced error and suggested that these flies were not monophyletic. Lastly, Linares et al. (2009) found Wolbachia infections i n their study of Madagascar butterflies, and showed that the use of generalized primers can enhance the amplification of Wolbachias DNA sequences, hence suggesting the parallel use of Wolbachia discovery methods coupled with mtDNA barcoding techniques. Nonetheless if the above issues prove difficult to control for in a study group, there are alternatives to COI mtDNA barcoding that have been developed and used to address the same goals. Hajibabaei et al. (2007) used DNA microarrays to identify and assess mammalian species, including felines, canines, bats, rodents, and primates among others, as explained previously Even though th e methodology is entirely unrelated to barcoding involving hybridization of sample DNA with DNA probes it has been shown to be useful for similar purposes. Furthermore, they also showed the functionality of the species level marker cytochrome b for the same goal. Similarly, alternative DNA regions have been used such as the ribosomal DNA (rDN A) Internal Transcribed spacer 2 (ITS2), which has a nuclear genome origin. Yao et al. (2010) suggested this molecular marker as a universal DNA region for plant and animal barcoding based on certain characteristics such as the possibility of developing universal primers t hat match its conserved regions. Much research has been done in this parallel field of DNA barcoding. It includes a broad array of taxonomic groups such as nematodes (Clapp et al., 2000) mites (Ben David et al., 2007) flatworms (Prasad et al., 2009) fig wasps (Li et al., 2010) and L ycaenidae butterflies (Wiemers et al., 2009) The latter work involves comparison of the COI mitochondrial marker and ITS2 nuclear


60 marker, in terms of their efficiency in species identification. Th e nuclear marker proved difficult to align (Wiemers et al., 2009) but inferring its secondary structure (i.e. translation of the genetic code into pseudoproteins) has proven to be effective when addressing lower taxonomic levels (Coleman, 2003) Wiemers et al. (2009) illustrate d the p erformance of both markers, and showed that both genes are comparable in their phylogenetic signal, thus advocating the use of a nuclear marker in part to overcome the inherent drawbacks that mtDNA barcoding presents (e.g. low mtDNA variation due to incomplete lineage sorting). Further studies have been done in Lepidoptera regarding the availability and utility of other markers for species level phylogenetics research. Wilson (2010) evalu ated the usefulness of the DNA barcode region in molecular systematics i n Lepidoptera in comparison with other markers including EF 1 WG, 18S, CAD, and DDC. He tested the performance of the COI barcode region at different taxonomic levels based on three criteria : i) practicality in terms of universality and sequence quality (reliability of the marker to produce sequences with low percentage of ambiguous bases) ii) phylogenetic utility regarding informative sites, and iii) accuracy of phylogenetic signal. For the first criterion, the COI region proved to be 100% successful in the 24 families surveyed as well as of high sequence quality for the same families. For the second criterion, the COI region showed good performance across taxonomic levels in relatio n to variable sites, parsimony informative sites, and minimum number of state changes, and was the most appropriate marker at lower taxonomic levels. The se results support the use of this marker in studies involving different genera, such as the present st udy of Preponini. Lastly, for the third criterion, COI was found to have similar


61 phylogenetic signal to nuclear markers at lower taxonomic levels. However the lowest category studied was genus, thus we lack information regarding the performance of COI in comparison with other markers at lower, species levels Using DNA barcoding effectively: the Prepona laertes case. Several outcomes are possible in using DNA barcoding to examine the taxonomy of the Prepona laertes group. Firstly, the nine different pheno types treated previously as species might prove to be reciprocally monophyletic, providing strong evidence that they are valid taxa. Alternatively, there may be no, or limited, differentiation among the phenotypes, consistent with the Lamas (2004) hypothes is of a single species. Most likely, however, some phenotypes will prove to be reciprocally monophyletic, others paraphyletic or polyphyletic, and still others undifferentiated. I aim to test the hypothesis proposed by Lamas (2004) in which Prepona laertes corresponds to one polymorphic species. Testing this hypothesis requires analyzing as many individuals and subspecies as possible, comprising several formerly recognized species such as P. philipponi and P. pseudomphale. Even though changes in coloration patterns can lead to speciation in butterflies through reproductive isolation (Jiggins et al., 2001; Jiggins et al., 2006; Lukhtanov et a l., 2005) an abrupt character change could be more likely to have an effect on reproductive isolation through, for instance, mate recognition. Still, hybridization can occur (Mallet et al., 2007) sometimes giving rise to new lineages (Mavarez et al., 2006) On the other hand, despite the fact that coloration changes can enhance reproductive isolation, sympatric speciation can arise with coloration stasi s (Jiggins et al., 2006) which suggests that not all variation in coloration implies speciation events. Therefore variation in coloration in butterflies may


62 not indicate speciation events but represent polymorphism, in which different m orphs can be found throughout a geographic distribution. Generally the characters that separate putative Prepona laertes group species are slight; they involve subtle changes from one putative species to other. Despite the fact that such wing pattern characters have been used to infer species, their presence might not imply speciation events, meaning that the variation seen can be attributed to intraspecific variation rather than to species specific features. For instance the main difference between P. pseudojoiceyi and P. joiceyi is the presence of a hook like termination of the light blue band on the forewing in P. pseudojoiceyi whereas in P. joiceyi the band ends abruptly (D'Abrera, 1987) Hence, such difference in coloration can be indicative of intraspecific variation instead of being a defining character of each species. If the Lamas hypothesis is correct, I predict all the individuals analyzed will have low divergence in the COI mitochondrial gene sequence, in comparison with inter specific divergence for other well defined Preponini species, independently of their phenotype. Additionally, I predict that individuals from the same putative species wil l not form clusters and/or will lack reciprocal monophyly (either for sympatric or allopatric species (Figure 32a)). On the other hand, clustering of barcode sequences would suggest the existence of additional species within Prepona laertes in which recip rocal monophyly of sympatric forms such as P. pseudomphale and P. philipponi would be strong evidence for ranking them as separate species (Figure 32b). If the clusters belong to allopatric forms (e.g. P. rothschildi and P. philipponi ), the amount of


63 dive rgence between the clusters should be quantified to decide whether the clusters merit species or subspecies status (Figure 32c). Materials and Methods Sample Acquisition and Origin As described in the previous chapter, in order to maximize the number of i ndividuals of Prepona laertes I carried out fieldwork in Colombia, and contacted collaborators in other countries, which provided tissue or COI sequence data (table 32). In addition to those samples, I searched for specimens in the collection of the McGui re Center for Lepidoptera and Biodiversity, as well as recently dead specimens from the Butterfly Rainforest at the Florida Museum of Natural History. Furthermore, I searched the GenBank ( ) repository to obtain sequences from other countries (table 3 2). DNA Extraction and Amplification DNA was extracted from two butterfly legs previously stored in 96% EtOH or recently removed from dried specimens. I used the Qiagen DNEasy Extraction Kit and followed their protocol. Howev er, some modifications were employed (see materials and methods in Chapter 2). In this case only the COI barcode region was amplified, following the protocol and primers discussed in the Materials and methods section of Chapter 2. DNA Sequencing and Alignm ent DNA sequencing from only one strand of the COI barcoding region was carried out at the University of Florida by the University of Floridas Interdisciplinary Center for Biotechnology Research Sanger Sequencing Group. I used Geneious v5.3 (Drummond et al., 2010) to edit and align the sequences. I constructed two matrices, one of 618bp


64 corresponding to 24 individuals that I processed. I created the second matrix after searching the Genbank database and complementing my dataset with additional sequences provided by D. Janzen (table 3 2). A total of 97 COI sequences composed this second matrix for a total length of 612bp. Cluster Analysis I used Neighbor Joining (NJ) (Saitou and Nei, 1987) to produce a cluster diagram in which the sequences in the dataset are grouped based on similarity (i.e. genetic distances). Genetic distances were estimated by the Kimura 2parameter model (K2P) (Kimura, 1980) with equal rates of variation. The outgroup species used in this analysis corresponds to COI sequence data for two individuals processed for the phylogenetic analysis, Prepona pylene eugenes H. W. Bates, 1865 and P. dexamenus dexamenus Hopffer, 1874. Results and Discussion The resulting to pologies corresponding to both datasets are shown in Figures 33 and 34. The COI sequence data I processed represented seven neotropical countries: Mexico, Honduras, El Salvador, Colombia, Venezuela, Ecuador and Brazil, for a total of 24 individuals. Despite the fact that all individuals analyzed belong to Prepona laertes according to Lamas (2004) they possess differences in their morphology, particularly in their coloration pattern. These characters have been used to define several species that currently are considered synonyms to Prepona laertes Phenotypes representing Prepona philipponi and P. pseudomphale, as defined by Neild (1996) were included in this analysis and are considered in both figures as different species to denote the differences in their morphology. Additionally, different geographical forms of Prepona laertes in reference to the Andes mountain range were represented in this analysis;


65 according to the Lamas taxonomy, these forms represent different subspecies, P. l. octavia Fruhstorfer, 1905 representing populations located to the west of the Andes and P. l. demodice (Godart, [1824]) to the east. Moreover, another morph is denoted in the analysis as Prepona laertes ssp1 corresponding to an unpubli shed classification of Colombian charaxines by J. F. LeCrom (Le Crom, pers. comm.). Yellow androconial tufts and the presence of light and dark blue dorsal bands characterize this putative new subspecies. The results show that there are no clear clusters of COI barcode sequences corresponding either to variations in coloration pattern (i.e. Prepona philipponi P. pseudomphale) or geographical origin in reference to the Andes (Figures 33, 3 4). Nonetheless there is a cluster of sequences regarding geograph ic region denoted by node 1. These sequences belong to individuals from Colombia and Venezuela and represent different morphological entities, including differences in androconial tufts coloration (orange in P. pseudomphale and yellow in P. laertes ) and pr esence/absence of dark blue band, for instance present in P. laertes ssp1 and absent in P. l. laertes (H bner, [1811]). Even though the origin of the samples is close (from the perspective of the entire geographic distribution represented in the analysis) this mixture of forms complicates the analysis since no correspondence is found between morphology and genetic data. Furthermore, these individuals are embedded in a polytomy within the outgroup species which obscures their real placement and their relation to the remaining barcode groups. Cluster number 2 represents individuals from diverse regions as well as of different phenotypes, while the most diverse cluster in terms of countries represented is


66 cluster 3. These two are well defined clusters with approximately 2% divergence. Nevertheless, this amount of genetic divergence is less than 3% suggested by Hebert et al. (2004b) to be a useful threshold to identify species. Hence, these clusters do not reach the amount of genetic divergence considered by some to be sufficient for species status, in addition to the fact that they do not correspond to sympatric groups with consistent morphological differences. The use of a fixed threshold of genetic divergence for recognizing species status within all groups has been widely cr iticized, and perhaps, these results would have provided evidence against their use if correspondence between morphological variations and genetic clusters had been found, coupled with a < 3% genetic divergence. An alternative approach to the use of fixed thresholds is comparing the divergence found for Prepona laertes clusters to the mean genetic divergence that characterizes other Prepona and Agrias species. I used the COI sequence data of nine Prepona and Agrias species included in the species level phy logeny to quantify the mean interspecific divergence and found that amount to be 9%. This suggests that the genetic divergence found for Prepona laertes is minimal compared to other preponine species. Then, I addressed the mean interspecific divergence between sister species of the Prepona and Agrias group and found that such divergence is 4.6 %, a result that further suggests that the genetic divergence seen in Prepona laertes is low in comparison to other preponine species. In terms of reciprocal monophyly, clusters 2 and 3 are reciprocally monophyletic, though they represent a wide variety of forms and a diverse geographic origin, which complicates the interpretation of the results since allopatric and sympatric forms are mixed together within the clusters. Similarly, there are two clear groups within cluster 3.


67 One corresponds to the Mexican + El Salvador samples, the other represents Colombian and Ecuadorian samples. Even though the Mexico + El Salvador cluster correspond to one morph, Prepona laertes octavia the second group represents different localities and contains different morphological variations, including P. l. octavia These clusters are also characterized by < 3% of genetic divergence. The results so far do not falsify the hypothesis proposed by Lamas (2004) in which he states that the former Prepona philipponi and P. pseudomphale are polymorphic forms of P. laertes However, the degree of divergence among individuals is also high, suggesting an alternative interpretation that multiple cryptic species lie within P. laertes that will only become clear through additional sampling. Figure 34 corresponds to all available COI data for Prepona laertes This analysis inc luded the data I processed plus additional samples from Mexico, Costa Rica and Panama. The Mexican samples are identified as Prepona laertes octavia according to the latest checklist (Lamas, 2004) Similarly, the Costa Rican samples are identified under the same subspecies, which according to D. Janzen (pers. comm.) should be considered two different species, which he denotes as Prepona demodice 1 and P. demodice 2 (a nomenclature not adopted here) His suggestion is based on ongoing research into not only COI barcode sequence data, but also natural history and ecology. Interestingly, there are two clusters that are mainly formed by Costa Rican samples, as noted by Janzen (Figure 34). However, ther e are several important issues to be addressed regarding my results. First, all Janzens samples come from one locality in Costa Rica, and such single region studies have prompted a debate regarding the minimum sufficient geographic breadth of the samples and level of sampling effort


68 (Brower, 2000; Meyer and Paulay, 2005) However, samples from other countries such as Ecuador, Panama, Honduras and Colombia also fall within the Costa Rican samples, showing that the Costa Rican clusters obtained by Janzen appear to remain stable with addition of samples from other countries, thus surpassing possible geographic barriers. Incomplete lineage sorting or retained ancient polymorphism might explain the pattern seen, so the inclusion of more individuals from throughout the geographic distribution could prove helpful in corroborating such apparent structure. Additionally, the Prepona laertes study could benefit from the inclusion of an additional marker aimed to test whether the COI sequence data are yielding accurate results. My results show that increasing geographic breadth might change the clustering pattern (Figure 3 4). Escalante et al. (2010) found Prepona laertes octavia to be monophyletic based on four COI sequences amplified from butterflies collected in the states of Guerrero and Ver acruz, but when I used those sequences in my analysis and included the additional Mexican sample I processed (collected in the state Tabasco), this monophyletic cluster disappeared. These results suggest that increasing the geographic dimension in the Prep ona laertes group indeed has an effect on the results, even though the geographic expansion was not apparently great (Tabasco state borders with Veracruz). Together, these results highlight the need for broad geographic sampling in future studies of Prepon a phylogeography. The most basal Prepona laertes samples form two clear clusters (denoted with numbers 1 and 2 in Figure 34) which are characterized by long branch lengths. Each cluster is represented by a diverse group of individuals, each possessing dif ferences in their coloration patterns (denoted in the figure as different species), as well as from


69 different geographical origins. Additional samples from as many localities and morphs should be included in the future in order to increase the COI sequence variation and help to clarify the results. For instance, including more samples from the same locality could be useful to reduce the observed long branch lengths, which might help resolve the relationships, for example, of the Colombian specimens (node 1 in Figures 33, 3 4). Including additional data is important, since the current results which indicate substantial differentiation might change with inclusion of additional individuals, showing apparent divergence to be continuous rather than discontinuous For example, in the case of Costa Rican specimens we would have observed a similar result as for the Colombian specimens if the analysis had only included a couple of specimens, such as those labeled with an asterisk. Based on the COI sequence data pres ented here, it is still unclear if there is a cryptic species complex within Prepona laertes. Despite the fact that some COI sequences form distinctive clusters (e.g. nodes 1 and 2 in Figure 3 4), the genetic distances in terms of branch lengths are high f or some individuals and negligible for others, which might potentially be an artifact of the lack of sampling. Thus, thorough geographic sampling and the inclusion of more individuals from the same localities could help disentangle the still unresolved tax onomy of the Prepona laertes group.


70 Figure 31. The DNA barcode gap. a. Presence of a barcode gap. b. Absence of a gap.


71 Figure 32. Possible outcomes for the Prepona laertes analysis. Blue curve corresponds to intraspecific comparisons of genetic dis tances of species 1, yellow line to intraspecific comparisons of genetic distances of species 2, and green line to interspecific comparisons of genetic distances for species 1 and 2. Only two putative species are shown for simplicity. a) Lack of barcoding gap and monophyletic clustering, b) Illustration of topology and barcode gap when sympatric forms are involved and c) when allopatric forms are involved.


72 Figure 33. NJ tree for the Prepona laertes complex. It incl udes 24 COI gene sequences from different putative species, which are currently considered within Prepona laertes such as P. philipponi and P. pseudomphale. Text in parenthesis corresponds to the country of origin as follows: MX: Mexico, HO: Honduras, EL : El Salvador, CO: Colombia, EC: Ecuador and BR: Brazil, and the location in relation to the Andes mountain range. Scale bar represents genetic distance. 1 2 3


73 Figure 34. NJ tree of all available data for Prepona laertes Text in parenthes is follows explanation in F igure 3 3 adding CR: Costa Rica and PA: Panama as a countries of sample origin. Scale bar represents genetic distance. 1 2


74 Figure 34. Continued.


75 Table 3 1. Historical classifications o f the tribe Preponini. Author Genus Prepona Agrias Archaeoprepona Anaeomorpha Noreppa Fruhstorfer (in Seitz 1907 1924) 14 7 10 1 n/a Vane Wright (1974) 9 n/a n/a n/a n/a Barselou (1983) n/a 4 n/a n/a n/a Spaeth (in D Abrera 1987) n/a 9 n/a n/a n/a D Abrera (1987) 21 n/a 5 1 2 Neild (1996) n/a 4 9 n/a 2 Lamas (2004) 7 5 8 1 1


76 Table 3 2. Specimen codes for COI sequence data used in the Prepona laertes study. Asterisk denotes Genbank accession numbers when available. Taxa Specimen Code Autho r Prepona laertes octavia HQ025040* Escalante et al. 2010 HQ025041* Escalante et al. 2010 HM888287* Escalante et al. 2010 HM888286* Escalante et al. 2010 HM416506* International Barcode of Life (iBOL), 2010 GU334298* Janzen and Hajibabaei, 2009 GU334299* Janzen and Hajibabaei, 2009 GU334300* Janzen and Hajibabaei, 2009 BCIBT154 09 Unpublished Janzen (iBOL) MHAAC001 06 MHAAC003 06 MHAAC004 06 MHAAC005 06 MHAAC006 06 MHAAC007 06 MHAAC008 06 MHAAC009 06 MHAAC010 06 MHAAC011 06 MHAAC012 06 MHAAC013 06 MHAAC014 06 MHAAC015 06 MHAAC016 06 MHAAC017 06 MHAAC018 06 MHAAC019 06 MHAAC020 06 MHAAC021 06 MHAAC022 06 MHAAC023 06 MHAAC024 06 MHAAC025 06 MHAAC026 06 MHAAC027 06 MHAAC028 06 MHAAC029 06


77 Table 3 2. Continued. Prepona laertes octavia MHAAC030 06 Unpublished Janzen (iBOL) MHAAC031 06 MHAAC032 06 MHAAC033 06 MHAAC034 06 MHAAC035 06 MHAAC036 06 MHAAC037 06 MHAAC038 06 MHAAC039 06 MHAAC040 06 MHAAC041 06 MHAAC042 06 MHAAC043 06 MHAAC044 06 MHAAC045 06 MHAAC046 06 MHAAC047 06 MHAAC048 06 MHAAC049 06 MHAAC050 06 MHAAC051 06 MHAAC326 07 MHAAC352 07 MHMXB880 06 MHMXG633 07 MHMXG634 07 M HMXG636 07 MHMXG643 07 MHMXS246 08 MHMXS257 08 MHMXY742 09 MHMXY743 09 MHMXY744 09 MHMYB035 09 MHMYB036 09


78 CHAPTER 4 CONCLUSIONS This phylogenetic study has suggested that several taxonomic modifications are required within the Preponini. They include modifications at different taxonomic levels, from tribal limits to species boundaries. More specifically, I recommend the exclusion of Anaeomorpha splendida from the tribe Preponini into an incertae sedis section. The genetic data used in this study provide strong evidence that Anaeomorpha is misplaced within the Preponini, but the true relationship of Anaeomorpha to other members of the subfamily Charaxinae is still inconclusive. The inclusion of additional markers and a broader t axonomic sampling would prove to be useful in determining the correct placement of Anaeomorpha, either as a member of an existing tribe, or perhaps as a new tribe. Regarding generic modifications, the genus Noreppa should be synonymized with Archaeoprepona syn. nov The combined results from genetic and morphological data (Marconato, 2008) support such merging, since the ML inferred tree shows that Noreppa chromus is embedded in the Archaeoprepona clade, and the synapomorphies used to define Archaeoprepona can also be used to identify Noreppa. I also suggest, according to the available data to date (i.e. genetic and morphological), that Agrias should probably be synonymized with Prepona despite the fact that other taxonomic solutions are possible. Although Agrias proved to be a well defined monophyletic group, it is placed well within the Prepona clade. Nonetheless, the bootstrap support and posterior probabilities are not particularly high. This leads me to conclude that additional markers would be required to yield more robust results before making a definite change, which could perhaps clarify Agrias closest relatives, thus providing more insight into the


79 relationships among members of this clade. In addition, I would encourage a detailed morphological study that attempts to identify synapomorphies that could perhaps support other taxonomic modifications (e.g. expanding Agrias to include P repona deiphile, P. werneri P. praeneste, and P. pylene). Also, I suggest that describing a new genus for the taxon pheridamas should be considerd, since, besides being basal to the Prepona + Agrias clade, it has distinctive morphology (i.e. coloration pattern and genitalia), and lacks morphological synapomorphies of remaining Prepona. In this case it would also be helpful to include an additional marker to increase the node support regarding Prepona pheridamas relationship to the remaining Prepona + Agrias species. This phylogenetic study of Preponini, besides allowing me to address my specific questions, has also indicated several taxonomic issues involving preponine species that require further study. It is clear that the relationshi ps among taxa included within Archaeoprepona amphimachus and A. meander deserve more research. Similarly, the phylogenetic relationships among the different subspecies of Prepona deiphile should be further investigated, since here they appear to be indepen dently evolving lineages and do not form a monophyletic group. In addition, the relationship between Agrias claudina and A. narcissus merits further research; the inclusion of more individuals representing different subspecies and from different geographic al origins might prove useful in defining their taxonomy. Likewise, broadening the taxonomic breadth of Prepona pylene samples would likely shed light on its true taxonomic status. Only two individuals were included in this analysis, both belonging to the same species according to Lamas (2004) but not to other researchers (e.g. Willmott, pers. comm.). Therefore, future study should include other forms of Prepona pylen e


80 Within the P repona laertes group, the hypothesis that Prepona laertes is a widespread polymorphic species as proposed by Lamas (2004) was not falsified by the COI barcode data. Even though dis tinctive clusters of COI barcode sequences were obtained, they did not show sufficient genetic divergence to consider them different species according to Hebert et al. (2003b) Additionally, when such divergence was compared to the mean genetic distances found for other preponine species, 9%, it is clear that the divergence found within Prepona laertes is low, even when addressing only the mean interspecific divergence between sister species (4.6 % for three species pairs). This conclusion is further supported by the lack of consistency in morphological characters within clusters. In spite of this, the tendency for clusters to have some geographic signal, such as the group form ed by Colombian and Venezuelan specimens, deserves further attention. The NJ trees constructed using the COI barcode data suggest that further sampling is required to better represent genetic variation and clarify results. My results also suggest that, contrary to what has been proposed previously, increasing the geographic scope of the study does appear to have an effect on the results. Therefore, I recommend thorough geographic sampling and inclusion of as many individuals as possible in future studies of the Prepona laertes group and of other Preponini species.


81 LIST OF REFERENCES Ackery, P.R., De Jong, R., VaneWright, R.I., 1999. The butterflies: Hedyloidea, Hesperioidea and Papilionoidea. In: Kristensen, N.P. (Ed.) Lepidoptera: Moths and Butterflies. 1. Evolutions, Systematics and Biogeography. Handbook of Zoology, 4(35), Lepidoptera. Berlin, De Gruyter, pp. 263300. Albrecht, A., Kaila, L., 1997. Variation of wing venation in Elachistidae (Lepidoptera: Gelechioidea): methodology and implications to systematics. Syst. Entomol. 22, 185198. Austin, G.T., 2009. Nymphalidae of Rondonia, Brazil: variation and phenology of Agrias (Charaxinae). Trop. Lepid. 19, 2934. Avise, J.C., Ball, R.M.J., 1990. Principles of genealogical concordance i n species concepts and biological taxonomy. Oxf. Surv. Evol. Biol. 7, 4567. Baker, R., Yu, X., Desalle, R., 1998. Assessing the relative contribution of molecular and morphological characters in simultaneous analysis trees. Mol. Phylogenet. Evol. 9, 4274 36. Barrett, R., Hebert, P., 2005. Identifying spiders through DNA barcodes. Can. J. Zool. 83, 481 491. Barselou, P., 1983. The Genus Agrias : a taxonomic and illustrated guide (Lepidoptera: Nymphalidae). Sciences Nat, Compigne. Ben David, T., Melamed, S., Gerson, U., Morin, S., 2007. ITS2 sequences as barcodes for identifying and analyzing spider mites (Acari: Tetranychidae). Exp. Appl. Acarol. 41, 169181. Bensasson, D., Zhang, D. X., Hartl, D.L., Hewitt, G.M., 2001. Mitochondrial pseudogenes: evolution's misplaced witnesses. Trends Ecol. Evol. 16, 314321. Bensasson, D., Zhang, D.X., Hewitt, G.M., 2000. Frequent assimilation of mitochondrial DNA by grasshopper nuclear genomes. Mol. Biol. Evol. 17, 406415. Berthier, S., 2005. Thermoregulation and spectral selectivity of the tropical butterfly Prepona meander : a remarkable example of temperature auto regulation. Appl. Phys. A Mater. 80, 13971400. Bickford, D., Lohman, D.J., Sodhi, N.S., Ng, P.K.L., Meier, R., Winker, K., Ingram, K.K., Das, I., 2007. Crypti c species as a window on diversity and conservation. Trends Ecol. Evol. 22, 148155. Bravo, A., Velez, P., Wolff, M., 2009. Mariposas del Parque Ecolgico Piedras Blancas. Fondo Editorial COMFENALCO. Medellin, Colombia.


82 Brito, P.H., Edwards, S.V., 2009. Mu ltilocus phylogeography and phylogenetics using sequence based markers. Genetica, 135, 439 455. Brower, A., 2006. Problems with DNA barcodes for species delimitation: ten species of Astraptes fulgerator reassessed (Lepidoptera: Hesperiidae). Syst. Biodiv ers. 4, 127132. Brower, A.V., 2000. Phylogenetic relationships among the Nymphalidae (Lepidoptera) inferred from partial sequences of the wingless gene. Proc. R. Soc. Lond. B 267, 12011211. Brower, A.V., 2010. Alleviating the taxonomic impediment of DNA barcoding and setting a bad precedent: names for ten species of Astraptes fulgerator (Lepidoptera: Hesperiidae: Eudaminae) with DNA based diagnoses. Syst. Biodivers. 8, 485491. Brownstein, J., Hett, E., O'Neill, S., 2003. The potential of virulent Wolb achia to modulate disease transmission by insects. J. Invertebr. Pathol. 84, 24 29. Chacn, I., Montero, J.J., 2007. Mariposas de Costa Rica. INBio Costa Rica. Clapp, J.P., van der Stoel, C.D., van der Putten, W.H., 2000. Rapid identification of cyst ( Hete rodera spp., Globodera spp.) and root knot ( Meloidogyne spp.) nematodes on the basis of ITS2 sequence variation detected by PCR single strand conformational polymorphism (PCR SSCP) in cultures and field samples. Mol. Ecol. 9, 1223 1232. Coleman, A.W., 2003. ITS2 is a doubleedged tool for eukaryote evolutionary comparisons. Trends Genet. 19, 370375. Combes, S., Daniel, T., 2003. Flexural stiffness in insect wings I. Scaling and the influence of wing venation. J. Exp. Biol. 206, 29792987. Cywinska, A., Hun ter, F.F., Hebert, P.D.N., 2006. Identifying Canadian mosquito species through DNA barcodes. Med. Vet. Entomol. 20, 413424. D'Abrera, B., 1987. Butterflies of the Neotropical Region. Part IV Nymphalidae (partim). Victoria, HillHouse. Darling, J.A., Blum, M.J., 2007. DNA based methods for monitoring invasive species: a review and prospectus. Biol. Invasions, 9, 751 765. Dedeine, F., Ahrens, M., Calcaterra, L., Shoemaker, D., 2005. Social parasitism in fire ants ( Solenopsis spp.): a potential mechanism for i nterspecies transfer of Wolbachia. Mol. Ecol. 14, 15431548. DeSalle, R., 2006. Species discovery versus species identification in DNA barcoding efforts: response to Rubinoff. Conserv. Biol. 20, 1545 1547.


83 DeSalle, R., Egan, M.G., Siddall, M., 2005. The unholy trinity: taxonomy, species delimitation and DNA barcoding. Philos. T. R. Soc. Lond. B, 360, 19051916. Descimon, H., 1977. Biog ographie, mim tisme et spciation dans le genre Agrias Doubleday (Lep. Nymphalidae Charaxinae). Publ. Lab. Zool. de l'Ecole. Norm Sup. 9, 307 344. Devries, P., Walla, T., 2001. Species diversity and community structure in neotropical fruit feeding butterflies. Biol. J. Linn. Soc. 74, 115. DeVries, P.J., 1987. The Butterflies of Costa Rica and Their Natural History Volume I: P apilionidae, Pieridae, Nymphalidae. Princeton University Press, New Jersey, U.S.A. Drummond, A.J., Ashton, B., Buxton, S., Cheung, M., Cooper, A., Heled, J., Kearse, M., Moir, R., Stones Havas, S., Sturrock, S., Thierer, T., Wilson, A., 2011. Geneious, Ver sion 5.4. Ekrem, T., Willassen, E., Stur, E., 2007. A comprehensive DNA sequence library is essential for identification with DNA barcodes. Mol. Phylogenet. Evol. 43, 530542. Elias, M., Hill, R.I., Willmott, K.R., Dasmahapatra, K.K., Brower, A.V.Z., Mall et, J., Jiggins, C.D., 2007. Limited performance of DNA barcoding in a diverse community of tropical butterflies. Proc. R. Soc. Lond. B, 274, 28812889. Escalante, P., Ibarra Vazquez, A., Rosas Escobar, P., 2010. Tropical montane nymphalids in Mexico: DNA barcodes reveal greater diversity. Mitochondr. DNA, 21, 3037. Felsenstein, J., 1985. Confidence limits on phylogenies: an approach using the bootstrap. Evolution, 39, 783791. Fral, J., 2002. How useful are the genetic markers in attempts to understand and manage marine biodiversity? J. Exp. Mar. Biol. Ecol. 268, 121145. Ferguson, J., 2002. On the use of genetic divergence for identifying species. Biol. J. Linn. Soc. 75, 509516. Folmer, O., Black, M., Hoeh, W., Lutz, R., Vrijenhoek, R., 1994. DNA primer s for amplification of mitochondrial cytochrome c oxidase subunit I from diverse metazoan invertebrates. Mol. Mar. Biol. Biotech. 3, 294 299. Freitas, A.V.L., Brown, K.S., 2004. Phylogeny of the Nymphalidae (Lepidoptera). Syst. Biol. 53, 363383. Fruhstorf er, H., 1916. Aegeronia, Brassolidae, Prepona, Morpho, Adelpha, Agrias In: Seitz (Ed.) The Macrolepidoptera of the World. Stuttgart, Alfred Kernen.


84 Funk, D., 1999. Molecular systematics of cytochrome oxidase I and 16S from Neochlamisus leaf beetles and the importance of sampling. Mol. Biol. Evol. 16, 67 82. Furtado, E., 2001. Prepona pheridamas pheridamas (Cramer) and its immature stages (Lepidoptera, Nymphalidae, Charaxinae). Rev. Bras. Zool. 18, 689694. Furtado, E., 2008. Intergeneric hybridism between Prepona and Agrias Trop. Lepid. 18, 5 6. Gaudeul, M., Taberlet, P., Till Bottraud, I., 2000. Genetic diversity in an endangered alpine plant, Eryngium alpinum L. (Apiaceae), inferred from amplified fragment length polymorphism markers. Mol. Ecol. 9, 16251637. Gompert, Z., Nice, C., Fordyce, J., Forister, M., Shapiro, A., 2006. Identifying units for conservation using molecular systematics: the cautionary tale of the Karner blue butterfly. Mol. Ecol. 15, 1759 1768. Gullan, P.J.a.P.C., 2004. The insects: an outline in entomology. Wiley Blackwell. U.K. Hajibabaei, M., Janzen, D., Burns, J., Hallwachs, W., Hebert, P., 2006. DNA barcodes distinguish species of tropical Lepidoptera. Proc. Natl Acad. Sci. U.S.A. 103, 968971. Hajibabaei, M., Singer, G.A.C., Clare E.L., Hebert, P.D.N., 2007. Design and applicability of DNA arrays and DNA barcodes in biodiversity monitoring. BMC Biol. 5, 24. Hall, B.G., 2008. Phylogenetic trees made easy. Sinauer Associates, Sunderland, U.K. Hall, J., Harvey, D., 2001. A phylogenet ic analysis of the Neotropical riodinid butterfly genera Juditha, Lemonias Thisbe and Uraneis with a revision of Juditha (Lepidoptera : Riodinidae : Nymphidiini). Syst. Entomol. 26, 453490. Heath, B., Butcher, R., Whitfield, W., Hubbard, S., 1999. Horiz ontal transfer of Wolbachia between phylogenetically distant insect species by a naturally occurring mechanism. Curr. Biol. 9, 313316. Hebert, P., Cywinska, A., Ball, S., DeWaard, J., 2003a. Biological identifications through DNA barcodes. Proc. R. Soc. Lond. B, 270, 313321. Hebert, P., Gregory, T., 2005. The promise of DNA barcoding for taxonomy. Syst. Biol 54, 852 859. Hebert, P., Penton, E., Burns, J., Janzen, D., Hallwachs, W., 2004a. Ten species in one: DNA barcoding reveals cryptic species in the neotropical skipper butterfly Astraptes fulgerator Proc. Natl Acad. Sci. U.S.A. 101, 14812 14817.


85 Hebert, P., Ratnasingham, S., deWaard, J., 2003b. Barcoding animal life: cytochrome c oxidase subunit 1 divergences among closely related species. Proc. R. Soc Lond. B, 270, S96S99. Hebert, P., Stoeckle, M., Zemlak, T., Francis, C., 2004b. Identification of birds through DNA barcodes. PLoS Biol. 2, e312. Hedges, L.M., Brownlie, J.C., O'Neill, S.L., Johnson, K.N., 2008. Wolbachia and virus protection in insects Science, 322, 702 702. Hillis, D.M., Moritz, C., Barbara, K., 1996. Molecular Systematics. Sinauer Associates. pp. 655. Hoy, M.A., 2003. Insect molecular genetics. Academic Press. San Diego, California, U.S.A. Huelsenbeck, J.P., Ronquist, F., Nielsen, R. Bollback, J.P., 2001. Bayesian inference of phylogeny and its impact on evolutionary biology. Science, 294, 23102314. Hundsdoerfer, A.K., Rubinoff, D., Attie, M., Wink, M., Kitching, I.J., 2009. A revised molecular phylogeny of the globally distributed hawkmoth genus Hyles (Lepidoptera: Sphingidae), based on mitochondrial and nuclear DNA sequences. Mol. Phylogenet. Evol. 52, 852865. Hurst, G., Jiggins, F., 2005. Problems with mitochondrial DNA as a marker in population, phylogeographic and phylogenetic studies: the effects of inherited symbionts. Proc. R. Soc. Lond. B, 272, 1525 1534. Iudica, C.A., Whitten, W.M., Williams, N.H., 2001. Small bones from dried mammal museum specimens as a reliable source of DNA. Biotechniques, 30, 732734. Janzen, D., Hajib abaei, M., Burns, J., Hallwachs, W., Remigio, E., Hebert, P., 2005. Wedding biodiversity inventory of a large and complex Lepidoptera fauna with DNA barcoding. Philos. T. Roy. Soc. B, 360, 18351845. Jiggins, C., Naisbit, R., Coe, R., Mallet, J., 2001. Rep roductive isolation caused by colour pattern mimicry. Nature, 411, 302305. Jiggins, C.D., Mallarino, R., Willmott, K.R., Bermingham, E., 2006. The phylogenetic pattern of speciation and wing pattern change in neotropical Ithomia butterflies (Lepidoptera : Nymphalidae). Evolution, 60, 14541466. Johnson, K., Descimon, H., 1988. Systematic status and distribution of the littleknown charaxinae Prepona werneri Hering & Hopp. J. Lepid. Soc. 42, 269 275. Johnson, K., Descimon, H., 1989. Proper generic and speci fic status of antillean Prepona butterflies with description of a new subspecies from Puerto Rico (Nymphalidae; Charaxinae). Caribb J. Sci. 25, 4553.


86 Joron, M., Wynne, I., Lamas, G., Mallet, J., 1999. Variable selection and the coexistence of multiple mim etic forms of the butterfly Heliconius numata. Evol. Ecol. 13, 721 754. Kaila, L., Stahls, G., 2006. DNA barcodes: Evaluating the potential of COI to diffentiate closely related species of Elachista (Lepidoptera : Gelechioidea : Elachistidae) from Australi a. Zootaxa, 126. Kandul, N., Lukhtanov, V., Dantchenko, A., Coleman, J., Sekercioglu, C., Haig, D., Pierce, N., 2004. Phylogeny of Agrodiaetus Hubner 1822 (Lepidoptera : Lycaenidae) inferred from mtDNA sequences of COI and COII and nuclear sequences of EF 1 alpha: karyotype diversification and species radiation. Syst. Biol. 53, 278298. Kimura, M., 1980. A simple method for estimating evolutionary rates of base substitutions through comparative studies of nucleotide sequences. J. Mol. Evol. 16, 111 120. Kon do, N., Nikoh, N., Ijichi, N., Shimada, M., Fukatsu, T., 2002. Genome fragment of Wolbachia endosymbiont transferred to X chromosome of host insect. Proc. Natl Acad. Sci. U.S.A. 99, 14280 14285. Lamas, G., 2004. Atlas of Neotropical Lepidoptera: Checklist Pt. 4a Hesperioidea Papilionoidea. Tropical Lepidoptera Scientific Publishers. Gainesville, Florida, U.S.A. Larkin, M.A., Blackshields, G., Brown, N.P., Chenna, R., McGettigan, P.A., McWilliam, H., Valentin, F., Wallace, I.M., Wilm, A., Lopez, R., Thompson J.D., Gibson, T.J., Higgins, D.G., 2007. Clustal W and Clustal X, Version 2.0. Bioinformatics, 23, 29472948. LeCrom, J.F., LlorenteBousquets, J., Constantino, L.M., Salazar, J., 2004. Mariposas de Colombia Tomo 2: Pieridae. Bogota, Colombia. Li, W., Go jobori, T., Nei, M., 1981. Pseudogenes as a paradigm of neutral evolution. Nature, 292, 237239. Li, Y., Zhou, X., Feng, G., Hu, H., Niu, L., Hebert, P.D.N., Huang, D., 2010. COI and ITS2 sequences delimit species, reveal cryptic taxa and host specificity of fig associated Sycophila (Hymenoptera, Eurytomidae). Mol. Ecol. Resour. 10, 3140. Linares, M., 1996. The genetics of the mimetic coloration in the butterfly Heliconius cydno weymeri J. Hered. 87, 142 149. Linares, M., Soto Caldern, I., Lees, D., Anth ony, N., 2009. High mitochondrial diversity in geographically widespread butterflies of Madagascar: a test of the DNA barcoding approach. Mol. Phylogenet. Evol. 50, 485495.


87 Little, D.P., Stevenson, D.W., 2007. A comparison of algorithms for the identifica tion of specimens using DNA barcodes: examples from gymnosperms. Cladistics, 23, 121. Liu, L., Pearl, D.K., Brumfield, R.T., Edwards, S.V., 2008. Estimating species trees using multiple allele DNA sequence data. Evolution, 62, 20802091. Llorente Bousquet s, J., Luis Martnez, A., Gonzlez Cota, L., 1992. Diferenciacin de Prepona deiphile en Mesoamrica y descripcin de dos subespecies nuevas (Lepidoptera: Nymphalidae). Trop. Lep. 3, 109114. Lukhtanov, V., Kandul, N., Plotkin, J., Dantchenko, A., Haig, D. Pierce, N., 2005. Reinforcement of pre zygotic isolation and karyotype evolution in Agrodiaetus butterflies. Nature, 436, 385389. Lukhtanov, V.A., Sourakov, A., Zakharov, E.V., Hebert, P.D.N., 2009. DNA barcoding Central Asian butterflies: increasing geographical dimension does not significantly reduce the success of species identification. Mol. Ecol. Resour, 9, 13021310. Magnacca, K.N., Brown, M.J.F., 2010. Mitochondrial heteroplasmy and DNA barcoding in Hawaiian Hylaeus ( Nesoprosopis ) bees (Hymenopter a: Colletidae). BMC Evol Biol 10, 174. Mallet, J., Beltran, M., Neukirchen, W., Linares, M., 2007. Natural hybridization in heliconiine butterflies: the species boundary as a continuum. BMC Evol. Biol. 7, 28. Marconato, G., 2008. Anlise Cladstica de Char axinae Guene (Lepidoptera, Nymphalidae). PhD Thesis. Instituto de Biociencias. Universidade de Sao Paulo. Mavarez, J., Salazar, C., Bermingham, E., Salcedo, C., Jiggins, C., Linares, M., 2006. Speciation by hybridization in Heliconius butterflies. Nature, 441, 868 871. Mayr, E., 1963. Animal species and evolution. Harvard University Press, Cambridge, U.K. Meier, R., Zhang, G., Ali, F., 2008. The use of mean instead of smallest interspecific distances exaggerates the size of the "barcoding gap" and leads to misidentification. Syst. Biol. 57, 809. Meyer, C., Paulay, G., 2005. DNA barcoding: error rates based on comprehensive sampling. PLoS Biol. 3, e422. Miller, J., Brower, A., Desalle, R., 1997. Phylogeny of the neotropical moth tribe Josiini (Notodontidae: Dioptinae): comparing and combining evidence from DNA sequences and morphology. Biol. J. Linn. Soc. 60, 297 316.


88 Monteiro, A., Pierce, N., 2001. Phylogeny of Bicyclus (Lepidoptera : Nymphalidae) inferred from COI, COII, and EF 1 alpha gene sequences. Mol. Phylogenet. Evol. 18, 264 281. Moritz, C., Cicero, C., 2004. DNA barcoding: promise and pitfalls. PLoS Biol. 2, e354. Muyshondt, A., 1974. Notes of the life cycle and natural history of butterflies of El Salvador. VIII. A rchaeoprepona antimache gulina, Sid erone marthesia, Zaretis callidryas and Consul electra (Nymphalidae). J. Lepid. Soc. 30, 159 168. Narita, S., Nomura, M., Kato, Y., Fukatsu, T., 2006. Genetic structure of sibling butterfly species affected by Wolbachia infection sweep: evolutionary and bi ogeographical implications. Mol. Ecol. 15, 10951108. Neild, A., 1996. The Butterflies of Venezuela. Meridian Publications, London. Pea, C., Wahlberg, N., 2008. Prehistorical climate change increased diversification of a group of butterflies. Biol. Letter s, 4, 274 278. Pea, C., Wahlberg, N., Weingartner, E., Kodandaramaiah, U., Nylin, S., Freitas, A.V.L., Brower, A.V.Z., 2006. Higher level phylogeny of Satyrinae butterflies (Lepidoptera: Nymphalidae) based on DNA sequence data. Mol. Phylogenet. Evol. 40, 2949. Petit, R., El Mousadik, A., Pons, O., 1998. Identifying populations for conservation on the basis of genetic markers. Conserv. Biol. 12, 844855. Posada, D., Crandall, K.A., 1998. Modeltest: testing the model of DNA substitution. Bioinformatics, 14, 817 818. Prasad, P.K., Tandon, V., Biswal, D.K., Goswami, L.M., Chatterjee, A., 2009. Use of sequence motifs as barcodes and secondary structures of internal transcribed spacer 2 (ITS2, rDNA) for identification of the Indian liver fluke, Fasciola (Trematoda: Fasciolidae). Bioinformation, 3, 314320. Prendini, L., 2005. Comment on" Identifying spiders through DNA barcodes". Can. J. Zool. 83, 498 504. Rambaut, A., 20062009. FigTree, Version 1.3.1. Rivera, J., Currie, D.C., 2009. Identification of Nearctic black flies using DNA barcodes (Diptera: Simuliidae). Mol. Ecol. Resour. 9, 224236. Rokas, A., Williams, B., King, N., Carroll, S., 2003. Genome scale approaches to resolving incongruence in molecular phylogenies. Nature, 425, 798804. Ronquist, F., Huels enbeck, J.P., 2003. MrBayes 3: Bayesian phylogenetic inference under mixed models. Bioinformatics, 19, 1572 1574.


89 Rubinoff, D., 2006. Utility of mitochondrial DNA barcodes in species conservation. Conserv. Biol. 20, 1026 1033. Rubinoff, D., Holland, B.S., 2005. Between two extremes: mitochondrial DNA is neither the panacea nor the nemesis of phylogenetic and taxonomic inference. Syst. Biol. 54, 952 961. Saitou, N., Nei, M., 1987. The Neighbor joining method: a new method for reconstructing phylogenetic trees. Mol. Biol. Evol. 4, 406 425. Salazar, J.A., 1999. Habitat e ilustracion del sintipo de Anaeomorpha splendida columbiana Niepelt, 1928 (Lep: Nymphalidae: Charaxinae) para Colombia. Bol. U. Cal. 29 32. Schwartz, M.K., Luikart, G., Waples, R.S., 2007. Gene tic monitoring as a promising tool for conservation and management. Trends Ecol. Evol. 22, 25 33. Shoemaker, D., Dyer, K., Ahrens, M., McAbee, K., Jaenike, J., 2004. Decreased diversity but increased substitution rate in host mtDNA as a consequence of Wolb achia endosymbiont infection. Genetics, 168, 20492058. Shoemaker, D., Keller, G., Ross, K., 2003. Effects of Wolbachia on mtDNA variation in two fire ant species. Mol. Ecol. 12, 1757 1771. Silva Brando, K.L., Lucci Freitas, A.V., Brower, A.V.Z., Solferin i, V.N., 2005. Phylogenetic relationships of the New World Troidini swallowtails (Lepidoptera: Papilionidae) based on COI, COII, and EF 1alpha genes. Mol. Phylogenet. Evol. 36, 468 483. Sites, J., Marshall, J., 2003. Delimiting species: a renaissance issue in systematic biology. Trends Ecol. Evol. 18, 462 470. Smith, M.A., Woodley, N.E., Janzen, D.H., Hallwachs, W., Hebert, P.D.N., 2006. DNA barcodes reveal cryptic host specificity within the presumed polyphagous members of a genus of parasitoid flies (Dipt era: Tachinidae). Proc. Natl Acad. Sci. U.S.A. 103, 3657 3662. 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. U.S.A. 105, 1348613491. Sorenson, M., Quinn, T., 1998. Numts: a challenge for avian systematics and population biology. Auk. 115, 214 221. Sota, T., Vogler, A., 2001. Incongruence of mitochondrial and nuclear gene trees in the carabid beet les Ohomopterus Syst. Biol. 50, 39 59.


90 Stamatakis, A., 2006. RAxML VI HPC: maximum likelihood phylogenetic analyses with thousands of taxa and mixed models. Bioinformatics, 22, 2688 2690. Stamatakis, A., Hoover, P., Rougemont, J., 2008. A fast bootstrappi ng algorithm for the RAxML webservers. Syst. Biol. 57, 758 771. Swofford, D.L., 2003. PAUP*. Phylogenetic Analysis Using Parsimony (* and Other Methods). Sinauer Associates, Sunderland, Massachusetts. Timmis, J.N., Ayliffe, M.A., Huang, C.Y., Martin, W., 2004. Endosymbiotic gene transfer: organelle genomes forge eukaryotic chromosomes. Nat. Rev. Genet. 5, 123135. Triplehorn, C.A., Johnson, N.F., 2005. Borror and Delong's Introduction to the Study of Insects. Brooks/Cole. pp. 888. VaneWright, R.I., 1974. Eugene LeMoult's Prepona types (Lepidoptera: Nymphalidae, Charaxinae). Bull. Allyn Mus. 21, 110. Wahlberg, N., Braby, M., Brower, A., de Jong, R., Lee, M., Nylin, S., Pierce, N., Sperling, F., Vila, R., Warren, A., Zakharov, E., 2005. Synergistic effects of combining morphological and molecular data in resolving the phylogeny of butterflies and skippers. Proc. R. Soc. Lond. B, 272, 15771586. Wahlberg, N., Leneveu, J., Kodandaramaiah, U., Pea, C., Nylin, S., Freitas, A.V.L., Brower, A.V.Z., 2009. Nymphali d butterflies diversify following near demise at the Cretaceous/Tertiary boundary. Proc. R. Soc. Lond. B, 276, 4295 4302. Wahlberg, N., Wheat, C.W., 2008. Genomic outposts serve the phylogenomic pioneers: designing novel nuclear markers for genomic DandNA extractions of lepidoptera. Syst. Biol. 57, 231 242. Werren, J., Windsor, D., Guo, L., 1995a. Distribution of Wolbachia among Neotropical Arthropods. Proc. R. Soc. Lond. B, 262, 197204. Werren, J., Zhang, W., Guo, L., 1995b. Evolution and phylogeny of Wol bachia: reproductive parasites of arthropods. Proc. R. Soc. Lond. B, Bio 261, 5563. Werren, J.H., Windsor, D.M., 2000. Wolbachia infection frequencies in insects: evidence of a global equilibrium? Proc. R. Soc. Lond. B, 267, 12771285. Whitworth, T.L., Da wson, R.D., Magalon, H., Baudry, E., 2007. DNA barcoding cannot reliably identify species of the blowfly genus Protocalliphora (Diptera : Calliphoridae). Proc. R. Soc. Lond. B, 274, 17311739. Wiemers, M., Fiedler, K., 2007. Does the DNA barcoding gap exis t? a case study in blue butterflies (Lepidoptera: Lycaenidae). Front. Zool. 4, 8.


91 Wiemers, M., Keller, A., Wolf, M., 2009. ITS2 secondary structure improves phylogeny estimation in a radiation of blue butterflies of the subgenus Agrodiaetus (Lepidoptera: Lycaenidae: Polyommatus). BMC Evol. Biol. 9, 300. Wiens, J.J., 2007. Species delimitation: new approaches for discovering diversity. Syst. Biol. 56, 875878. Wiens, J.J.a.T.A.P., 2002. Delimiting species using DNA and morphological variation and discordant species limits in spiny lizards ( Sceloporus ). Syst. Biol. 51, 69 91. Will, K., Mishler, B., Wheeler, Q., 2005. The perils of DNA barcoding and the need for integrative taxonomy. Syst. Biol. 54, 844851. Will, K., Rubinoff, D., 2004. Myth of the molecule: D NA barcodes for species cannot replace morphology for identification and classification. Cladistics, 20, 4755. Wilson, J.J., 2010. Assessing the value of DNA barcodes and other priority gene regions for molecular phylogenetics of Lepidoptera. PLoS ONE, 5, e10525. Yao, H., Song, J., Liu, C., Luo, K., Han, J., Li, Y., Pang, X., Xu, H., Zhu, Y., Xiao, P., Chen, S., 2010. Use of ITS2 region as the universal DNA barcode for plants and animals. PLoS ONE, 5, e13102. Zhou, W., Rousset, F., O'Neil, S., 1998. Phylog eny and PCR based classification of Wolbachia strains using wsp gene sequences. Proc. R. Soc. Lond. B, 265, 509515.


92 BIOGRAPHICAL SKETCH Elena Ortiz Acevedo is a young biologist broadly interested in ecology and evolution, using diurnal butterflies as a model to understand the mechanisms that generate and maintain species diversity. She is particularly interested in understanding the causes of speciation and diversification, and as a consequence she has lately focused her work in phylogenetics and specie s delimitation. She graduated in 2007 from Universidad de los Andes in Bogot, Colombia, where butterflies first attracted her attention during her undergraduate thesis. For this work, she compared butterfly diversity in two riparian forests in the eastern foothills of the eastern cordillera in Colombia. She worked in the field over several months, after which her interest in butterfly diversity led her to continue in the pursuit of knowledge about Colombian butterflies. After graduating, she worked for the Tropical Andean Butterfly Diversity Project, digitizing the information available in the butterfly collection of the Instituto de Genetica at Universidad de los Andes, allowing her to increase her knowledge about the diversity and distribution of Colombi an butterflies, as well as improving her skills in managing large amounts of data and information. She volunteered for the museum for natural history at Universidad de los Andes where she discovered her passion for biological collections and learned about curatorial techniques. She has also written several grants, of which one of the first was funded by the Tropical Andean Diversity Project to compare butterfly communities in certified and noncertified shade coffee plantations. She executed this project fr om field design to data analysis, and a resulting paper will be submitted in 2011 to the journal Conservation Biology. In addition to these two projects, she has attended two workshops and field courses about butterflies. At the first


93 course she attended as a student, and at the second she was invited as an organizer, as a result of her experience in organizing logistics and knowledge about Colombian butterflies. In 2009, she became a masters student at the Department of Entomology and Nematology and McGu ire Center for Lepidoptera and Biodiversity, at the Florida Museum of Natural History at the University of Florida. There, she became interested in the butterfly tribe Preponini, an interest which drove her to propose and execute a project in which she aim ed to resolve the phylogenetic relationships of the tribe and clarify the number of species in the group. During her masters, she additionally worked as a Research Assistant in the McGuire Centers butterfly collections, where she further improved her knowledge about neotropical butterflies and also curatorial techniques of biological collections. These experiences have maintained her interest in butterfly research, improved her skills in planning and executing research projects, enhanced her abilities for critical thinking and data analysis, and finally have increased her commitment to working with biological collections, a commitment which she hopes to keep developing for many years.