Citation
Phylogenetic studies of the Melitaeini (Lepidoptera: Nymphalidae: Nymphalinae) and a revision of the genus Chlosyne Butler

Material Information

Title:
Phylogenetic studies of the Melitaeini (Lepidoptera: Nymphalidae: Nymphalinae) and a revision of the genus Chlosyne Butler
Creator:
Kons, Hugo L., 1974-
Publication Date:
Language:
English
Physical Description:
x, 798 leaves : ill. ; 29 cm.

Subjects

Subjects / Keywords:
Colors ( jstor )
Cream ( jstor )
Hair ( jstor )
Marinas ( jstor )
Phenotypes ( jstor )
Phylogenetics ( jstor )
Species ( jstor )
Symmetry ( jstor )
Taxa ( jstor )
Term weighting ( jstor )
Cladistic analysis ( lcsh )
Dissertations, Academic -- Entomology and Nematology -- UF ( lcsh )
Entomology and Nematology thesis, Ph. D ( lcsh )
Nymphalidae -- Phylogeny ( lcsh )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 2000.
Bibliography:
Includes bibliographical references (leaves 790-796).
Additional Physical Form:
Also available online.
General Note:
Printout.
General Note:
Vita.
Statement of Responsibility:
by Hugo L. Kons, Jr.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright [name of dissertation author]. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
025854725 ( ALEPH )
46727538 ( OCLC )

Downloads

This item has the following downloads:


Full Text










PHYLOGENETIC STUDIES OF THE MELITAEINI (LEPIDOPTERA: NYMPHALIDAE: NYMPHALINAE) AND A REVISION OF THE GENUS
CHLOSYNE BUTLER













By

HUGO L. KONS, JR.














A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE
UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA 2000









Dedicated to the collectors of Lepidoptera:



An extensive amount of work crucial to the success of this project was completed before it even began, in the form of the extensive collections of specimens available for study. These specimens were collected, curated, and labeled from all over the world, and represent the cumulative efforts of many amateur and professional Lepidopterists collecting in many states and nations over a period exceeding 100 years. Many of the Melitaeini are strikingly beautiful butterflies long popular among collectors, and their beauty combined with diurnal habits have contributed to their representation in collections. Nevertheless, some taxa remain poorly represented, and more material would have been highly desirable. Many other groups of Lepidoptera, let alone other insects, are far more poorly represented in collections today.

In the face of the expanding human population and ever increasing destruction of natural habitats, and consequent crisis for understanding and preserving Earth's biodiversity, the need for collecting and systematic studies of insects has never been greater. Yet, tragically, amateur and professional Lepidopterists in many lands presently have their efforts stifled by a gauntlet of misguided collecting restrictions based on ignorance, territorialism, and/or financial greed, including: restrictions preventing the acquisition of material for study, restrictions preventing collectors from retaining the specimens they collect, restrictions on transporting dead specimens across political lines, an overburden of unnecessary red tape and/or timely delays associated with obtaining permits, restrictions which discriminate against amateur Lepidopterists and/or people outside particular political boundaries, and/or exorbitant fees required to purchase permits even for noncommercial scientific collecting. It is a sad tragedy at a time when collecting is needed more then ever, it is nonetheless discouraged more than ever, thus severely limiting the prospects of success for any state, national, or international effort to understand, document, and preserve biodiversity.

In recognition of the indispensable role of collectors to the study and preservation of

Lepidoptera biodiversity, I dedicate this work to the collectors of Lepidoptera, past, present, and future, amateur and professional, and institutional and private, in the hope that the gauntlet of misguided regulations so plaguing the efforts of Lepidopterists around the world will undergo repeal or sweeping reform in a future time of far greater enlightenment and cooperation.















ACKNOWLEDGMENTS

I wish to acknowledge a number of people who provided valuable assistance with this project. For serving on my doctoral committee, providing advice and assistance throughout this project, and for review of this manuscript, I thank Thomas Emmel, James Lloyd, Jonathen Reiskind, Frank Slansky, and John Heppner. For providing access to or loan of specimens critical to this study I thank John Heppner (Florida State Collection of Arthropods), Robert Robbins and Donald Harvey (National Museum of Natural History), Lee and Jackie Miller (Allyn Museum), Keith Willmott, and Jason Hall. I am grateful to Gerardo Lamas for sending me scanned images of several Chiosyne types in the British Museum. For providing work space in the Florida State Collection of Arthropods I thank John Heppner. Thomas Walker provided access to an exceptional camera lucida microscope set up for doing the genitalia illustrations. I thank Steve Lasley and Nick Hostettler for computer assistance. I also thank Thomas Emmel and the University of Florida's Department of Entomology and Nemnatology (especially the graduate coordinator, Grover Smart) for providing research and teaching assistantship funding, respectively, for some semesters. Donald Harvey and Robert Robbins provided considerable advice and assistance during a visit to the National Museum of Natural History, and Robert Robbins assisted with securing a short term visitors grant for me to visit that institution. Debbie Hall was most helpful for assistance dealing with the University of Florida's bureaucracy. For helpful discussions and/or debates on species problems, cladistics, and/or other issues in taxonomy I thank James Lloyd, Donald



ill









Harvey, Robert Robbins, Jonathen Reiskind, Keith Willmott, Jason Hall, Byron Adams, Vitor Becker, Clay Scherer, J. Akers Pence, and Paul Choate. Finally, I thank my parents, Hugo and Sharon Kons, Sr., for encouragement throughout my time as a graduate student.











































iv
















TABLE OF CONTENTS

pane
ACKNOWLEDGMENTS.............................................................................11i

ABSTRACT ......................................................................................... ix

CHAPTERS

IINTRODUCTION...............................................................................1.

Overview .........................................................................................1.
Natural Taxa, Testable Hypotheses, and a Universal Taxon Characteristic ................... 3
Species and Subspecies Concepts Used in this Work: Theoretical Concepts of the
Species/subspecies Category and Practical Applications for Delimitating
Species/subspecies Taxa ......................................................................... 7
The Species Problem ...................................................................... 7
The Theoretical Species Concept Used in this Work .................................. 9
Applying the Evolutionary Species Concept in Practice to Delimitating Species
and Subspecies of Butterflies .....................................................11..I
Subspecies Concepts Rejected in this Work........................................... 13

2 PHYLOGENETIC ANALYSIS OF THE MELITAEIM (LEPIDOPTERA: NYMPHALIDAE: NYMPHALINAE) AND A REVISION OF THE HIGHER CLASSIFICATION OF THE MELITAE1NI......................................................... 16

Introduction.......................................................................................... 16
Materials and Methods
Out Groups................................................................................ 19
Type of Characters Coded for Analysis................................................ 23
In Group Taxa Examined .............................. :................................. 24
Preparation of Melitaeini Genitalia for Character Analysis........................ 26
Production of Melitaeini Genitalia Figures to Illustrate Character States ......... 27 Phylogenetic Analysis ................................................................... 27
Results............................................................................................. 28
Characters and Character States for a Phylogenetic Analysis of the Melitaeini
(Lepidoptera: Nymphalidae)......................................................... 28
Phylogenetic Analysis ................................................................... 82
Discussion ........................................................................................... 86
Revised Higher Classification of the Melitaeini ................................................. 95
Melitaeini Tutt ........................................................................... 96


v









Euphydryiti H iggins ........................................................................................... 98
M elitaeiti T utt ....................................................................................................... 105
G nathotrichiti Subtribe N ..................................................................................... 107
Phycioditi H iggins ................................................................................................. 114
Poladryiti Subtribe N ........................................................................................... 116
C hlosyniti Subtribe N ........................................................................................... 12 1
Key to the Subtribes of the Melitaeini Based on Genitalic Characters .................... 123

3 PHYLOGENETIC STUDIES OF THE CHLOSNYITI AND POLADRYITI (LEPIDOPTERA: NYMPHALIDAE: NYMPHALINAE: MELITAEINI) .............................. 213

In troduction .......................................................................................................................... 2 13
M aterials and M ethods ......................................................................................................... 215
M aterial E xam ined ................................................................................................ 2 15
C haracter C oding .................................................................................................. 2 18
Phylogenetic Analyses Including the Entire Data Set ........................................... 219
Investigation of the Impact of Different Numbers of Taxa in the Analysis on Tree
and C haracter Statistics .................................................................................... 222
Test for Congruence Between Genitalic and Pattern Characters .......................... 223
Test of the Impact of Different Numbers of Taxa on Boot Strap Scores ........... 224
R esu lts .................................................................................................................................. 2 24
Character Coding Overview .................................................................................. 224
Characters and Character States for a Phylogenetic Analysis of the Chlosyniti and
Poladryiti Based on Morphological Data ......................................................... 226
PhylogeneticAnalyses Based on the Entire Data Set ............................................ 332
Variation in Tree Statistics with Different Numbers of Taxa in the Analysis ...... 338
Variation in Character Statistics with Different Numbers of Taxa in the
A n aly sis ............................................................................................................ 339
Variation in the Proportion of Homoplastic Characters with Different Numbers of
T axa in the A nalysis ......................................................................................... 340
Congruency Between Tree Topologies Derived from Genitalic and Pattern
C h aracters ......................................................................................................... 34 1
Variation in Boot Strap Scores for Various Clades as a Result of the Number of
Taxa Included in the Analysis .......................................................................... 343
D iscu ssio n ............................................................................................................................ 34 4
Proposed Hypothesis for the Phylogeny of Chlosyniti and Poladryiti .................. 344
Systematic Check List of the Chlosyniti ............................................................... 344
Phylogenetically Invalid or Unsupported Generic Concepts ................................ 346
Phylogenetically Invalid or Unsupported Species Concepts ................................ 350
Different Models for Character State Polymorphisms .......................................... 357
Variation in Tree and Character Statistics with Different Numbers of Taxa ....... 365
Variation in Boot Strap Scores with Different Numbers of Taxa in the
A n aly sis ............................................................................................................ 36 7
Equal Character Weighting, Successive Character Weighting, and Successive
Character State Weighting: An Alternative Method Proposed for Generating
O ptim al T rees ................................................................................................... 372


vi









Congruency Between Tree Topologies Derived From Genitalic and Pattern
Characters ......................................................................................................... 382

4 PHYLOGENETIC REVISION AND MORPHOLOGICAL CHARACTERIZATION OF THE CHLOSYNITI (LEPIDOPTERA: NYMPHALIDAE: NYMPHALINAE: M ELITAEIN I) ........................................................................................................................... 439

Introduction .......................................................................................................................... 439
M aterials and M ethods ......................................................................................................... 440
Key to the Genera, Species, and Subspecies of the Chlosyniti ................................ 444
Antillea Higgins ................................................................................................................... 459
Antillea proclea (Doubleday and Hewitson) ........................................................ 464
Antillea pelops (Drury) ......................................................................................... 466
The M icrotia Bates and Chlosyne Butler Clade .................................................................. 468
M icrotia Bates ..................................................................................................................... 469
M icrolia eleda (Hew itson) .................................................................................... 472
M icrotia elva Bates ............................................................................................... 478
M icrotia dymas (Edwards) .................................................................................... 483
M icrotia coracara (Dyar) ..................................................................................... 486
M icrotia anomalus (Godm an & Salvin) ............................................................... 490
Chlosyne Butler .................................................................................................................... 491
Chlosyne harrisii (Scudder) .................................................................................. 495
Chlosyne kendalloruin Opler ................................................................................ 507
Chlosyne nycteis (Doubleday) .............................................................................. 510
Chlosyne gorgone (Hubner) .................................................................................. 516
Chlosyne hoffmanni .............................................................................................. 521
Chlosynepalla (Boisduval) ................................................................................... 529
Chlosyne whitneyi (Skinner) ................................................................................. 536
Chlosyne gabbii (Behr) ......................................................................................... 543
Chlosyne acastus (W H Edwards) ...................................................................... 549
Chlosyne definita (Aaron) ..................................................................................... 556
Chlosyne ezra (Hew itson) ..................................................................................... 564
Chlosyne chinaliensis (Tinkham ) .......................................................................... 570
Chlosyneperlula (M enetries) ............................................................................... 578
Chlosyne theona (M enetries) ................................................................................ 586
Chlosyne leanira (Felder & Felder) ...................................................................... 598
Chlosyne endeis (Godm an & Salvin) .................................................................... 632
Chlosyne marina (Geyer) ...................................................................................... 636
Chlosyne melitaeoides (Felder & Felder) ............................................................. 644
Chlosyne erodyle (Bates) ...................................................................................... 647
Chlosyne inelanarge (Bates) ................................................................................. 653
Chlosyne eumeda (Godm an & Salvin) .................................................................. 656
Chlosyne hylaeus (Godm an & Salvin) .................................................................. 659
Chlosyne californica (W right) .............................................................................. 661
Chlosyne lacinia (Geyer) ...................................................................................... 667
Chlosyne ehrenbergi (Geyer) ................................................................................ 685


vii









Chlosyne hippodrome (Geyer) .............................................................................. 689
Chlosyne narva (Fabricius) ................................................................................... 694
Chlosyne gaudealis (Bates) ................................................................................... 699
Chlosynejanais (Drury) ........................................................................................ 705
Chlosyne rosita Hall .............................................................................................. 716

5 SUM M ARY AND CONCLUDING REM ARKS ............................................................... 757

APPENDIX PLATES OF ADULT CHLOSYNITI & POLADRYITI .................................. 763

REFERENCES .......................................................................................................................... 790

BIOGRAPHICAL SKETCH ..................................................................................................... 797









































viii















Abstract of Dissertation Presented to the Graduate School of the University of Florida in
Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy

PHYLOGENETIC STUDIES OF THE MELITAEINI (LEPIDOPTERA:
NYMPHALIDAE: NYMPHALINAE) AND A REVISION OF THE GENUS CHLOSYNE BUTLER

By

Hugo L. Kons, Jr.

December 2000

Chair: Thomas C. Emmel
Major Department: Entomology and Nematology

An evolutionary hypothesis is presented for the phylogenetic relationships within the Melitaeini (Lepidoptera: Nymphalidae: Nymphalinae), based on cladistic analyses of morphological characters. Melitaeini is revised at the subtribal level, and the concepts of Euphydryiti and Phycioditi but not Melitaeiti are upheld as natural groupings. To create a natural classification scheme, Melitaeiti is restricted to a clade of Eurasian taxa, and three new subtribes are proposed for taxa formerly placed in Melitaeiti: Chlosyniti, Poladryiti, and Gnathotrichiti. Species level phylogenetic analyses are presented for all subtribes except the Melitaeiti and Phycioditi. Euphydryiti is revised at the generic level, and includes the genera Euphydryas Scudder, Hypodryas Higgins, and Eurodryas Higgins, with Occidryas Higgins synonymized with Euphydryas. Gnathotrichiti is revised at the generic level and includes a single monophyletic genus, Gnathotriche Felder and Felder, with Gnathotrusia Higgins placed in synonymy. Poladryiti is revised




ix









at the generic level, and includes three monophyletic genera: Atlantea Higgins, Higginsius Hemming, and Poladryas Bauer. Chlosyniti is revised in detail at the generic and species levels, and includes three monophyletic genera: Antillea Higgins, Microtia Bates, and Chlosyne Butler, with the genera Dymasia Higgins, Texola Higgins, Charidryas Scudder, Thessalia Scudder, and Anemaca placed into synonymy. Thirtyseven species taxa and seventeen subspecies taxa are recognized within the Chlosyniti. Detailed camera lucida drawings illustrate all genitalic characters and character states. The phylogenetic analysis of Poladryiti and Chlosyniti is used as a case study to investigate some issues of broader implication to systematic biology, including equally versus successively weighting characters, tree statistics, homoplasy, boot strap scores, and polymorphisms. Separate analyses of genitalic and pattern characters were conducted for the Cholsyniti/Poladryiti data matrix to investigate the proportion of groupings in conflict between independent data sets for different methods of analysis. The percentage of conflicting groupings was 0% for equal weighting of characters with parsimony, 36.8% for successive weighting and parsimony, 77% for a phenetic algorithm (UPGMA), and 100% between pairs of random trees. The absence of incongruence between independent data sets for the former analysis is argued as evidence supporting the effectiveness and theoretical validity of the phylogenetic methods used in this study.














x














CHAPTER 1
INTRODUCTION


Overview

The Melitaeini (Lepidoptera: Nymphalidae: Nymphalinae) include a group of over 265 butterfly species with representatives in Nearctic, Palearctic, and Neotropical regions with the greatest diversification in the last (Higgins 1960, Higgins 1981). The New World representatives of this group have been the subject of several major revisions, including Higgins (1960), Higgins (1978), and Higgins (1981). The classifications resulting from these revisions were based primarily on similarities and differences in adult morphology. However, previous to the present work, the Melitaeini had never been investigated with the tools of cladistic analyses to construct an evolutionary hypothesis of the natural relationships within the group.

My initial interest in the Melitaeini was based on an interest in conducting a

phylogenetic study of the genus Chlosyne Butler and its close relatives. What attracted me to this genus included a great complexity in morphological variation associated with pattern characters (an impression that was immediately evident examining specimens in museum drawers), and as a strikingly patterned group of butterflies long popular among collectors, the genus was particularly well represented in collections. The combination of complex morphological variation and extensive material available for study led me to believe (correctly) that the Chlosyne were a group well suited for pursuing my interests in phylogenetic studies of Lepidoptera. Early in the project, I decided that to achieve my



I






2

goals with respect to a phylogenetic study of the genus Chlosyne, a study of the higher relationships within the Melitaeini as a whole was desirable, and I expanded my interest to include a phylogenetic study of the tribe. While I studied only a sample of the taxa which Higgins placed in the Phycioditi and of the Palearctic fauna which Higgins (1981) placed in the Melitaeti (designed only to test if these groupings were monophyletic and if so to determine their placement on the phylogenetic tree), I was able to include almost all other melitaeinine taxa in the investigations covered in this work.

The major objectives of this work can be broadly defined as follows: 1) To

determine whether the Melitaeini comprise a natural group (Chapter 2); 2) To derive an evolutionary hypothesis of the major clades within the Melitaeini and of their relationships to each other (Chapter 2); 3) To determine which taxa are most closely related to the genus Chlosyne Butler, and then to derive an evolutionary hypothesis for the relationships among these taxa (Chapter 3); 4) To use the genus Chlosyne and its relatives as a case study to investigate some issues of broader interest to the field of phylogenetic systematics (Chapter 3); 5) To modify existing classification schemes of the Melitaeini, if necessary, based on the criteria of monophyly (=creating a natural classification scheme) and stability (Chapters 2 to 3), and to characterize the natural higher groupings of taxa morphologically; and 6) To revise and morphologically characterize Chlosyne and its relatives, recognizing species and subspecies taxa based on phylogenetic, morphological, distributional, and if available, biological evidence (Chapter 4).

I divide this work into three major chapters. The first (Chapter 2) is a study of the higher relationships within the Melitaeini, and of the relationships within the species






3


comprising several of the smaller clades, based on morphological characters coded from genitalic structures. The second (Chapter 3) is a detailed phylogenetic study of Chiosyne and its relatives, and uses this group as a case study to investigate a number of issues of interest to phylogenetic systematics. The final major chapter (Chapter 4) utilizes the knowledge obtained from conducting the studies in Chapter 3, combined with other information, to revise and morphologically characterize Chiosyne and its close relatives.

An overriding theme that I hope prevails throughout this work is systematics as a scientific discipline based upon testable hypotheses of true patterns that exist in nature. It has been my goal to study and analyze the morphology of the Melitaeini with the best scientific tools available to the modern systematist, and to part ways to the greatest extent possible with arbitrary decision-making which has been prevalent in systematics in the past. Particulars of this goal are presented in all of the three major chapters, but in the introduction to the entire work, I find it appropriate to address two issues of general significance: 1) What constitutes a valid taxon in my vision of biological systematics; and 2) The complexities of the species problem, and how the modern tools of systematics can be used in actual practice to delimit natural species and subspecies taxa within Lepidoptera (in particular), based on testable hypotheses.


Natural Taxa, Testable Hypotheses, and a Universal Taxon Characteristic

In a review of the previous literature regarding the classification of the Melitacini, I find quotes such as the following to be plentiful: "these differences are so marked that I have not found it possible to write a generic synopsis to include both..." (Higgins 1960), "The specialised characters [of CharidryasJ do not appear to be sufficiently marked to justify generic separation" (Higgins 1960), "the species associated with Charidrv as ... are






4


sufficiently distinct in maculation to merit recognition" (Ferris 1989), "Higgins included three Mexican species in the genus, but the other two species are about as close to Microtia and Dymasia as they are to the type species, Texo1a eleda" (Bauer 1975), and "The dark populations in southern Mexico have been referred to as the nominotypical subspecies but are sufficiently different to recognize" (Austin and Smith 1998b). An overriding theme that I find in common with these and many similar statements regarding taxa delimitation in the literature is arbitrary decisions being based on a worker's perception of similarity and difference. These statements and many others like them likewise illustrate the conflict and instability that results from classifications based on such criteria, as different workers often have differing views on what is most similar, dissimilar, or sufficiently different to warrant recognition. What I find absent in the above quotes, is a testable hypothesis upon which taxon delimitation is based, and by which future workers objectively may analyze evidence to determine if the taxa delimited are valid. Furthermore, of what relevance to the goals of biological systematics are my personal perceptions (or those of any other worker) regarding what entities are most similar or different?. While accepted under the International Code of Zoological Nomenclature, I personally reject the criteria of sufficient similarity/sufficient difference as a basis for delimitating taxa of any rank in the scientific discipline of biological systematics because 1) Such criteria represent arbitrary opinions rather than testable hypotheses (regardless of a workers level of experience); and (2) Such criteria often result in the delimitation of artificial taxa.

I define a natural taxon as a group of organisms representing a true product of the natural process of evolution with a real existence in nature, as opposed to an artificial






5

taxon, which represents a product of the human imagination with no real existence in nature. In other words, natural taxa are monophyletic evolutionary lineages, including by definition all the descendents of a common ancestor. Artificial taxa are any groupings of organisms which a human being artificially creates, and do not represent an evolutionary lineage. I make no distinction between paraphyletic and polyphyletic taxa. A paraphyletic group has been defined (originally by W. Hennig) as a group not including all of the descendents of a common ancestor (Scotland 1992a), while a polyphyletic group has been defined as "a group in which the most recent common ancestor is assigned to some other group and not to the group itself' (Scotland 1992a, quoted from E. Wiley via J. Farris). Because all life is believed (based on current evidence) to have a common origin, any conceivable grouping of extant organisms that is not monophyletic is by definition paraphyletic. Consequently, for the purposes of classifying extant organisms in a natural classification scheme, the only true distinction is between monophyletic (natural) and nonmonophyletic (artificial) groups.

Artificial classification schemes can be very useful for certain purposes, such as deciding what to include in a book, what to include in a university course, what to include in a pest control manual, etc. Some examples of paraphyletic groups sharing some type of similarity used for these purposes include reptiles, invertebrates, moths (excluding butterflies), the Lepidoptera of Florida, Florida pests of citrus, medically important insects, urban pests, beautiful moths, etc. Although I recognize the utility of such groupings for these purposes, I argue that such groupings (or any nonmonophyletic grouping) have no place being formally named or ranked in the scientific discipline of biological systematics for several reasons: 1) They are artificial, do not exist in nature,






6


and lack evolutionary significance; 2) The scientific method cannot be used to test their validity, as they have no characteristic corresponding to a testable hypothesis in common with other groups; 3) They are a poor fit with a hierarchical classification of sets within sets and nonoverlapping sets used in systematic classification; and 4) They would result in highly unstable classifications, because there are many different ways people can view organisms as being similar or different.

As the study of real patterns in nature and the use of testable hypotheses are an obvious part of any scientific discipline, if systematics is to be viewed as a scientific discipline the importance of the former two reasons is evident. With regard to the third reason, systematic classification does not allow, for example, one species to be a member of two different genera. If one were to ask the question, is Taxon A more similar to Taxon B or to Taxon C, different workers could easily arrive at different answers because there are many ways that taxa can be viewed as similar or different, yet Taxon A cannot be placed in two different genera even if it shares some type of similarities with both. However, there is only one way that Taxa A, B, and C can be related in terms of evolutionary relationships.

Although the hierarchical classification of systematics was developed before the theory of evolution, I argue that this hierarchy nonetheless lends itself well to a natural classification and relatively poorly to an artificial one. The decision that Taxon A is more similar to B than C is ultimately an arbitrary opinion, and any other worker could just as validly arrive at the opposite opinion. The decision that Taxon A is more closely related to B than C is a testable hypothesis. In order to change Taxon A's placement with Taxon B in a natural classification, one would need to provide additional evidence






7


showing that the initial hypothesis is no longer favored, as opposed to arbitrarily deciding that Taxon A is more similar to Taxon C. Even in a natural classification, disputes can still arise over the validity of particular taxa as two workers may reach conflicting hypotheses from the same evidence. However, the scientific method is self-correcting over time, as new evidence and/or an improved theoretical framework favors one hypothesis over its alternatives.

In all aspects of this work, I choose the criterion of monophyly as a universal requirement for the validity of a taxon of any rank. While monophyly provides a scientific basis for testing the validity of taxonomic concepts, it does not address another issue in the systematic classification of organisms: how many natural taxa does one formally name and how are they ranked. Above the species level, I see no way such a question can be answered with the scientific method, and the issue is a matter of bookkeeping rather than science. However, I argue that making such bookkeeping decisions objectively is preferable to arbitrarily doing so. This goal can be achieved by applying a secondary criterion of nomenclatural stability, as discussed in Chapters 2 and 3 where phylogenetic evidence is used to convert the existing classification schemes into a natural classification scheme.


Species and Subspecies Concepts Used in this Work: Theoretical Concepts of the
Species/subspecies Category and Practical Applications for Delimitating Species/subspecies Taxa

The Species Problem:

Ereshefsky (1992) makes an important distinction regarding the use of the term it species." One usage is the species category, relating to what characteristics make an entity a species, and the other usage is species taxa, that which is formally named as a






8


species by taxonomists. Biologists have widely differing views on how the species category is defined (Ereshefsky 1992), and consequently species taxa are delimited in a variety of different ways. In my view, a clear-cut universally applicable species concept is unachievable, both in theory and practice. If species are required to be natural (=monophyletic) lineages, the speciation process in its simplest special case involves a dichotomous branching event where an original population of one species diverges over time into two species. Bell (1997) points out that the unit event of the evolutionary process is an episode of variation (=reproduction in sexually reproducing organisms) followed by an episode of selection. It seems untenable that between two such episodes there would occur a clear transformation of an entity changing from one species to a different species, or from a subspecies to a species, or from a population to a subspecies. Whatever characteristics are assigned to the species category or a practical method of delimitating species taxa, because evolution is an ongoing process, as an ancestral lineage begins to diverge into two lineages (by a series of episodes of reproduction followed by episodes of selection) that may eventually become clear-cut separate species, there will be a period in-between when the two divergent lineages are in the process of acquiring the characteristics that make them species. The status of such lineages will always represent a gray area in delimitation of species taxa. Also, while it is clear that organic diversity falls into clusters (separate lineages) as opposed to a continuum of genetic variation (Ereshefsky 1992), Mishler and Donoghue (1992) point out that units corresponding to these clusters may not be directly comparable, and that discontinuities between clusters (such as morphological versus reproductive) may not correspond. I would infer from my studies of the Chlosyniti that no two clusters of organic diversity






9


that I recognize as basal taxa are equivalent units, because divergence between clusters varies in kind and degree. In my view, it is essential that taxonomists first objectively describe the patterns of variation that they find in nature, and then translate this information into species/subspecies decisions, as opposed to basing descriptions of the patterns of variation in nature on an attempt to conform to a particular species/subspecies concept.


The Theoretical Species Concept Used in this Work

Although I find the goal of a universally applicable clear-cut species concept

untenable, I do conclude that one of the species concepts proposed has the greatest utility to the field of biological systematics. In my view, a species concept useful to biological systematics should have two important characteristics: 1) All species units are natural (=monophyletic) taxa; and 2) Species units have an additional biologically and evolutionarily significant characteristic that distinguishes them from higher and lower monophyletic groups, such as genera and populations, respectively. However, with any species concept one must just accept that there are some cases where lineages will be in a state of transition in the speciation process, and in these cases species taxa delimitations will have to be made more arbitrarily.

The species concept that in my view best achieves the above goals is the

evolutionary species concept described by Wiley (1992), where species are defined by: "A species is a single lineage of ancestral descendent populations of organisms which maintains its identity from other such lineages and which has its own evolutionary tendencies and historical fate" (Wiley 1992). 1 regard a more concise definition as: a species is a monophyletic lineage biologically incapable of reticulating with a different






10


evolutionary lineage. The point where an evolutionary lineage loses its ability to merge with another lineage is in my view the theoretically and biologically significant event that separates a species lineage from a lower lineage. Although Wiley (1992) argued that a major advantage of his species concept was its applicability to all types of organisms, with respect to Lepidoptera (sexually reproducing organisms) a comparison can be made with the Biological Species Concept of Mayr (1963), where species were defined as "Groups of actually or potentially interbreeding populations, which are reproductively isolated from other such groups." For sexually reproducing organisms, the criteria of 11not actually or potentially interbreeding" and "maintains its identity from other such lineages" are very similar, with the exception of "gray area" cases where some amount of interbreeding occurs between two lineages yet they remain somewhat distinct. While reproductively isolated lineages may remain very similar morphologically over long periods of time, as exemplified by morphologically very similar butterflies and moths isolated on different mountain ranges, for two separate lineages to reticulate into one, reproduction must occur. In fact, with respect to sexually reproducing organisms, in my view the evolutionary species concept is a consequence of putting the biological species concept into an evolutionary context, achieved by adding a requirement that species units be natural products of the evolutionary process. A very serious flaw of the biological species concept is that it may recognize artificial (=nonmonophyletic) taxa not useful for phylogenetic studies when applied to allopatric populations (Mishler and Donoghue 1992). Regarding allopatric populations, O'Hara (1994) points out a flaw common to both the biological and evolutionary species concepts. In some situations, the decision of whether two lineages are separate species is based on knowing what will happen in the






11

future (will circumstances arise in the future where two currently allopatric lineages will come together and be capable of loosing their distinct identities or interbreeding?). Applying the Evolutionary Species Concept in Practice to Delimitating Species and Subspecies of Butterflies:

As a place to start, I search for gaps in the range of morphological variation of one or more characters. Assemblages of individuals that I find have a consistent morphological discontinuity become basal taxonomic units to be entered into the phylogenetic analysis. The presence of a consistent morphological discontinuity represents a testable hypothesis by which to initially delimit terminal taxa. Of course, this gap criterion is not a perfect one, but it does provide a place to start. It is possible for two distinct lineages to exist within a group of individuals that I cannot distinguish morphologically. Hypothetically, those lineages may not even be sister lineages. It is also possible that gaps between groups of taxa could be the result of polymorphism in the same lineage (C. hylaeus versus C. eumeda? See Chapter 4), or a result of insufficient collecting to show a true geographic continuum that exists in nature (C. rosita mazarum versus C. rosita riobalensis See Chapter 4). An advantage of selecting a relatively well collected group to study, such as Chlosyne, is that the risk associated with these two types gap errors is minimized. The monophyly of terminal Operational Taxonomic Units (OTUs) identified by morphological discontinuity need not be hypothesized based on the gap alone. Additional evidence to support the monophyly of terminal OTUs can be achieved by entering them into the phylogenetic analysis twice and seeing if they come out as sister taxa (supporting monophyly) or in a polytomy (providing evidence neither for nor against monophyly).






12


The testable hypotheses on which species delimitation are based include the

following: 1) Is the species taxon monophyletic?; and 2) is the species taxon biologically capable of reticulating with another evolutionary lineage (present or future)?. The first hypothesis is tested by cladistic analysis. Evidence that I find that substantially supports (although not necessarily guarantees) the second hypothesis includes any one or a combination of the following: 1) Sympatric occurrence of two distinct taxa; 2) Discontinuity in the range of morphological variation of a sclerotized structure, especially in the structure of the genitalia; 3) A lineage has no sister species, and is sympatric with at least one of the taxa in its sister dlade; or 4) Evidence of hybrid inviability with closely related lineages. In these cases, the evolutionary species concept is applied easily in actual practice, because the evidence supporting the testable hypotheses of an evolutionary species is fairly clear.

However, now consider a monophyletic group of two allopatric taxa with identical genitalia, no known information regarding hybrid inviability, but with morphological discontinuities in wing pattern or other pattern characters. Based on this evidence, these taxa may be hypothesized as distinct evolutionary lineages, but it is impossible to know whether or not they are evolutionary species. This is where I feel a subspecies concept can be useful. Theoretically, I define the subspecies category as follows: A subspecies is a single lineage of ancestral descendent populations of organisms that currently maintains its identity from other such lineages but that has not lost its biological ability to reticulate with another lineage except by geographic isolation. The future fate of a subspecies could be a reticulation event with another subspecies or further divergence into a separate species. In practical application, I delimit subspecies






13


taxa based on the presence of evidence to suggest they are distinct lineages but the absence of evidence to suggest they are evolutionary species. In other words, my definition of a lepidopteran subspecies in practical application is: a monophyletic lineage, allopatric with its closest relatives, with which it has identical genitalia (and other scierotized structures) but discontinuities in wing pattern. A subspecies could be increased in rank to a species if evidence is acquired that favors the hypothesis tested to delimit an evolutionary species over an alternative hypothesis.


Subspecies Concepts Rejected in this Work:

The vast majority of Chlosyne subspecies included by various authors do not meet either my criteria for a valid taxon or a valid subspecies. A common method of delimiting subspecies within the Chlosyniti has been to name populations, groups of populations, or groups of individuals comprising parts of different populations, representing points along a continuum of geographic variation. Geographic gradients are often non-uniform, with phenotypes remaining more constant over some areas and changing more rapidly over others. Subspecific status is assigned to populations occupying various geographic areas based on the criterion (extremely arbitrary in my view) that the worker(s) naming the subspecies feel it is sufficiently distinct to warrant recognition (see quote by Austin and Smith 1 998a above). Populations that represent intermediates between the named populations are not assigned to any subspecies taxon, and are referred to as subspecies A/B blends or intermediate populations (see Austin and Smith 1 998a, Austin and Smith 1 998b) or as unassigned or undetermined populations (see Smith and Brock 1988). Because some populations or individuals within the species are omitted from the subspecies concepts, the subspecies concepts are by definition






14


artificial. I reject all such subspecies as artificial taxa delimitated by arbitrary decisions rather than natural taxa delimitated by testable hypotheses.

I also do not view such subspecies concepts as useful for describing geographic variation. Consider the character of forewing length which varies geographically in C. theona (Austin and Smith 1998b). Austin and Smith (1998b) report the mean measurements and range (low-high) for a variety of artificial subspecies taxa, but do not report such information for populations designated as "blend populations". If the goal is to describe geographic variation, the blend populations arejust as significant as any population within the artificial subspecies concept, and may be of particular interest if they represent areas of higher variability or a steeper phenotypic gradient. For the purposes of quantifying geographic variation, I argue the best approach would have been to measure individual populations and to present this data, as opposed to presenting measurements for an artificial assemblage of populations.

The second type of subspecies concept which I reject is naming populations within a continuum of variation but with reported different averages in the range of variation for a character compared with adjacent populations. I distinguish between population differences and subspecies differences. Because selective pressures are unlikely ever to be completely identical between two separate populations, invariably every local population has a difference in the average range of variation for some character compared with adjacent populations. If such differences provide a basis for naming subspecies, every local population, no matter how temporary, would have to be named as a subspecies. If one requires that the average difference be sufficiently distinct






15


to warrant subspecies recognition, once again taxa are being delimitated by arbitrary opinions as opposed to testable hypotheses.















CHAPTER 2
PHYLOGENETIC ANALYSIS OF THE MELITAEINI (LEPIDOPTERA:
NYMPHALIDAE: NYMPHALINAE) AND A REVISION OF THE HIGHER CLASSIFICATION OF THE MELITAEINI


Introduction

The higher classification of Nymphalidae that I adopted as a starting point is the one proposed by Harvey (1991). Harvey's (1991) concept of Nymphalidae includes approximately 6,452 species worldwide (Shields 1989), and thirteen subfamilies including Heliconiinae, Nymphalinae, Limenitidinae, Charaxinae, Apaturinae, Morphinae, Brassolinae, Satyrinae, Calinaginae, Danainae, Tellervinae, Ithominae, and Libytheinae. Previous classifications, including Ehrlich (1958), Miller and Brown (1981), and Ackery (1988), ranked Melitaeinae as a subfamily of a narrower concept of Nymphalidae, with some of Harvey's (1991) subfamilies ranked as separate families. Higgins (1981) included three tribes within Melitaeinae, the Melitaeini, Phyciodini, and Euphydryini. A consequence of Harvey's (1991) classification is that Melitaeinae is given a tribal rank and the three tribes of Higgins (1981) are now ranked as subtribes Melitaeiti, Phycioditi, and Euphydryiti.

Harvey (1991) recognized three tribes within the Nymphalinae, including

Nymphalini, Kallimini, and Melitaeini. The synapomorphy that Harvey (1991) used to define this group was the larval character of a scolus on abdominal segment 9 (A9) ventral and posterior to the filiform setae. Harvey (1991) noted that Heliconiinae also




16






17


have a scolus on A9, but because it occurs in a different position (dorsal to the filiform setae), he presumed that it was not homologous to that of the Nymphalinae.

Harvey (1991) reported that systematic relationships within the Nymphalinae were poorly understood, particularly at the generic and tribal levels. The relationships beamong Nymphalini, Kallimini, and Melitaeini are unknown. Nymphalini and Melitaeini share a feature unique to the Nymphalidae of having filiform setae present on A 1,2, while Kallimini and Melitaeini share a feature unique to the Melitaeini of having the filiform setae on A9 occurring on the sclerotized base of the scolus (Harvey 1991). Consequently, at least one of these characters must have been subject either to reversal, or potentially less likely, to independent acquisition.

As noted by Harvey (1991), Higgins (1981) never provided a synapomorphy for the Melitaeini. Harvey (1991) reported the notched saccus in the male genitalia as a unique character for Melitaeini among the Nymphalidae, but actually several Phycioditi lack a notched saccus, so this character does not represent a universal synapomorphy. Harvey (1991) also noted that the distribution of filiform setae on Melitaeinine larvae is diagnostic within the Nymphalidae, based on the combination of filiform setae on A1,2 and on A9 at the sclerotized base of the scolus. However, although the combination of these features is a good identification character, it provides no evidence of monophyly for the Melitaeini, because each is present in some out group taxa.

The following is a summary of Higgins' (1981) higher classification of the

Melitaeini, which includes three subtribes (ranked as tribes by Higgins (1981)) and 31 genera. The species that Higgins (1981) included in each genus appear on pages 165-171 of his 1981 review of the classification of the Melitaeini.






18




Higgins' (1981) higher classification of the Melitaeini (with tribes down-ranked to subtribes):

Euphydryiti Higgins: Euphydryas Scudder, Hypodryas Higgins, Occidryas Higgins, and Eurodryas Higgins

Melitaeiti Tutt: Mellicta Billberg, Melitaea Fabricius, Poladryas Bauer, Didymaeformia Verity, Cinclidia Hubner, Chlosyne Butler, Thessalia Scudder, Texola Higgins, Dymasia Higgins, Microtia Bates, Gnathotriche Felder & Felder, Gnathotrusia Higgins, Higginsius Hemming, Antillea Higgins

Phycioditi Higgins: Phyciodes Hubner, Phystis Higgins, Anthanassa Scudder, Dagon Higgins, Telenassa Higgins, Ortilia Higgins, Tisona Higgins, Eresia Boisduval, Castilia Higgins, Janatella Higgins, Mazia Higgins Unplaced Genus: Atlantea Higgins


The support of a final phylogenetic tree can be no stronger than the support of any assumption made along the way, and because phylogenetic trees generated in parsimony analysis with morphological data are rooted and the characters are then polarized based on the out group (Kitching 1992a), the out group decision is critical. Although the best evidence currently available suggests that Nymphalinae forms a monophyletic group including Nymphalini, Kallimini, and Melitaeini (Harvey 1991), the relationships among these three tribes are unclear (Harvey 1991). Consequently the closest relatives of the Melitaeini are unknown. A similar problem of out group uncertainty was addressed with respect to a phylogenetic study of Rekoa Kaye (Lycaenidae: Theclinae) by Robbins (1991). Robbins (1991) presented evidence that Rekoa was related to Arawacus Kaye and Thereus Johnson, but was unsure which represented the sister group and unsure of the of relationships within these two genera. Robbins (1991) solved this problem by including taxa in these two genera as the out group taxon, and by coding them






19


collectively as one out group taxon. This out group taxon was coded for a state not present in the in group (=9) for characters not useful for tree rooting because one of the following was the case: 1) Out group character information was missing; 2) all in group character states occurred in the out group; or 3) None of the in group character states occurred in the out group. For the remaining characters, Robbins (1991) coded the out group taxon as the state or combination of states occurring in the in group that also occurred in the out group.

The objectives of this study were to code morphological characters for cladistic analysis to 1) Test the hypothesis that the Melitaeini form a monophyletic group; (2) To reconstruct a phylogenetic hypothesis of relationships within the Melitaeini; and (3) To improve the characterization of the Melitaeini based on adult morphology. Objective 2 included testing the validity of the subtribes and genera recognized by Higgins (1981 ) (with the exception of those within the Phycioditi and Eurasian Melitaeiti, for which the aim of the study was only to determine whether those groups were monophyletic and the placement of their clades on the Melitaini phylogeny), and to revise the higher classification of the Melitaeini such that natural evolutionary relationships are reflected by the higher taxa recognized. Finally, a major aim of this study was to determine which genera are related to Chlosyne, and to determine an out group for an additional detailed phylogenetic study of Chlosyne.


Materials and Methods


Out Groups






20


1 examined a number of taxa within Nymphalini and Kallimini to try to avoid making unwarranted assumptions on character polarization based on missing data. I adopt a similar out group approach to that of Robbins (199 1) for combining Nymphalini and Kallimini as the out group for a phylogenetic study of the Melitaeini, with my coding guidelines outlined below. Some of my coding decisions are different from those of Robbins (1991) in order to provide more specific information in the matrix, such as distinguishing among the three scenarios where Robbins coded the out group "9". However, their affect on the phylogenetic analysis is unchanged. For the purposes of discussion, I refer to this method as the "Cumulative Out Group Method."

I designate "c" as the symbol for a state occurring in out group taxa that never

occurs within the in group, and "n" and "p" as different character states that occur in both the in group and out group (in the actual data matrix "n" or "p" would be designated as a numerical state, except for multistate characters with more than 10 states where state assignments follow the sequence "0,1,2,3,4,5,6,7,8,9 ,a,b,d" (letters are used because MacClade 3.07 interprets "10,' as "0 & 1"). The following is a summary of how I coded characters for the cumulative out group: 1) All out group taxa exhibit a state (n) which also occurs in the in group: the out group is coded "n". 2) Some out group taxa have state n, while others have state p, and both states n and p occur within the in group: the out group is coded "n,p" (=n&p in MacClade) (if no states occur within the in group other than n and p, coding the out group as "T' or "c" would be equivalent in terms of affecting the computer analysis, but less informative regarding the distribution of character states among out group taxa). The same coding methodology would apply to characters with three or more states occurring in both the out group and in group. 3) No






21


out group taxon exhibits a state that occurs in the in group: the out group is coded "c" (the number of states occurring in out group taxa, or how they would be designated is irrelevant to cumulative out group coding). 4) Some out group taxa exhibit a state that does not occur in the in group, while others exhibit one or more states that do occur in the in group (n and p): the out group is coded "c,n,p." (5). Some out group taxa are coded for a particular character, but state information is unknown for one or more taxa included in the cumulative out group: the out group is coded "?". 6) A character clearly falls into discrete states (n and p) within the in group, but not among out group taxa: the out group is coded "?", while in group taxa are individually coded "n" or "p" as appropriate.

In an attempt to include a sample representative of the character state variation within Nymphalini and Kallimini, I examined the males of the type species of each Nymphalini and Kallimini genus listed in Harvey's (1991) Table B.2 except for a few genera for which I did not have males of the type species available (Bassaris, Symbrenthia, and Mynes). Information on the specific types for these genera was obtained from Hemming (1967). I also included a representative of Liminetidini, Colobura dirce, in the out group. The taxa comprising the cumulative out group are listed in Table 1 under "Out Group Taxa". If the cumulative out group method is applied to the phylogenetic analysis of a genus where several genera are potentially sister taxa, for scoring the cumulative out group it would be best to include all species taxa in the genera which are potential sister taxa. However, with respect to the phylogenetic analysis of a tribe such as Melitaeini with two subtribes as potential sister taxa, an approach of examining all species taxa in Kallimini and Nymphalini would be prohibitively impractical.






22


With respect to the cumulative out group method, I have polarized characters within the Melitaeini based on two assumptions. The first is that the sister group to Melitaeini includes and is exclusively composed of some combination of taxa currently classified within the Nymphalini and Kallimini. The second assumption is that the sample of generic types is representative of the character states that occur both within the true Melitaeinine out group and the in group (Melitaeini). I have made no assumption that particular taxa or groupings of taxa within the Nymphalini and Kallimini are most closely related to Melitaeini.

In addition to the use of the cumulative out group method, I developed a second method for polarizing characters within the Melitaeini (in a separate analysis) which I term the "Representative Out Group Method." Use of the representative out group method requires that characters are first scored for the cumulative out group method. From the resulting data matrix, I then selected representative out group taxa, such that the group of representative taxa collectively includes all character states occurring within the in group. These taxa were then entered into the data matrix individually rather than collectively, and specific states were assigned for each taxon in place of the "c" designation (invalid for this analysis because "c" can refer to more than one character state). I scored some additional characters that are invariant within Melitaeini but for which there is variation in states among the representative out group taxa. The reason for doing this was so that all the representative out group taxa would not, evidence permitting, come out as a unresolved basal polytomy where the number of equally parsimonious out group permutations would be multiplied by the number of equally






23

parsimonious trees within the in group, thereby greatly increasing the time needed to do an analysis in PAUP.


Type of Characters Coded for Analysis

The data matrix obtained for the Melitaeini is based exclusively on genitalic

characters. I searched for other external characters of sclerotized structures, but did not find variation that could be coded into discrete states within the in group. Scale pattern characters were not used in this analysis, because I did not have series of material from throughout the geographic range for a portion of the taxa examined (wing pattern can be highly geographically variable within the Melitaeini, but with a few nonproblematic exceptions I have found almost no geographic variation in genitalia). Another reason for not using pattern characters in this analysis is that many taxa in the Phycioditi and Eurasian Melitaeiti were not included in the analysis, and while the samples examined may provide a very good indication of the range of variation in relevant genitalic characters (=characters informative with respect to the monophyly of Phycioditi and Eurasian Melitaeiti and the placement of their clades on the phylogeny, as opposed to only informative with respect to relationships within these groups), they may not with respect to wing pattern.

Homologous characters existing in more than two forms among the taxa

examined were coded as multistate characters as opposed to multiple binary characters. Coding of certain complex multistate characters is discussed on a character by character basis in the "Results: Characters and character states for a phylogenetic analysis of the Melitaeini based on genitalic characters" section below. The general guidelines I strived to follow when making character coding decisions include: (1) no a priori assumptions






24

are made regarding character state order, (2) character coding schemes which weight the same change more than once are avoided, (3) coding schemes which may produce unwarranted bias regarding character state polarity are avoided, (4) a continuum of variation is not arbitrarily divided up into a series of discrete states, (5) characters which seem to be completely dependent (such as two separate wing patches always being the same color-more relevant to the following chapter) are coded only once, (6) for characters involving further division of a character state into additional states which do not apply to taxa lacking the state to be divided, "?" coding is avoided if but only if this can be accomplished without weighting an identical step twice, and (7) characters which can be coded into discrete states are not deleted because a priori I feel they may be homoplastic.


In Group Taxa Examined

Taxa examined and coded for characters are listed in Table 1. All described Melitaeini taxa other than Phycioditi and Eurasian Melitaeiti were included with the following exceptions for which I did not have material available for study: Eurodryas alexandria (Staudinger), Eurodryas orientalis (Herrich-Schafter), Gnathotrusia steinii (Dewitz), and Atlantea cryptadia. Also, I did not dissect the few specimens I examined of Atlantea perezi (Herrich-Schaffer); however, judging from figures 11-13 (pg. 391) in Higgins (1960), I suspect there would be very few if any differences in character coding between this taxon and the other Atlantea examined. Several taxa had to be coded for males only, since I had no females available for study. These taxa are Texola anomalus (Godman & Salvin), Higginsiusfasciatus (Hopffer), Gnathotriche sodialis Staudinger, and Gnathotrusia mundina (Druce).






25

Since the Phycioditi and Eurasian Melitaeiti are large groups, with 137 and 67

species, respectively, listed in Higgins (1981), and since the aim of this study was only to determine if these groups are monophyletic and where their clade appears on the Melitaeinine phylogeny, I limited my coverage of these groups to a sample. I coded characters for 18 Phycioditi, including the type species of each genus. I also examined Higgins (1981) Figures 183-477, which include genitalia drawings of males of most of the Phycioditi, and while these figures are limited in their detail, they do suggest the sample of taxa examined provides good representation of the range of genitalic variation within Phycioditi. I suspect most or all characters and character states not included in the sample are particular to certain Phycioditi, and consequently not important to an analysis aiming only to determine if the Phycioditi are monophyletic and the position of their clade on the Melitaeinine tree. I coded characters for 13 species of Eurasian Melitaeiti, including the types of each genus plus several additional representatives.

The number of male individuals which I dissected varied among the in group taxa. For taxa examined in the genera Chlosyne, Thessalia, Charidryas, Anemaca, Texola, Dymasia, Microtia, Antillea, and Poladryas, I dissected a minimum of three individuals from different parts of a taxon's range, and in nearly all cases more than three individuals were examined. The sole exception is Texola anomalus, for which I was only able to examine one specimen. For Phycioditi and Eurasian Melitaeini, two or more individuals were examined for the type species of each genus, and one to two individuals for the remaining taxa. For Euphydryiti, three or more individuals were dissected for Nearctic taxa and the types of each genus, while one to two individuals were dissected for remaining taxa. I dissected only one individual of Gnathotriche sodialis and






26


Gnathotrusia mundina, two individuals of Atlantea tulita, and three or more individuals of the remaining taxa. The vesica was everted, or partially everted, for at least one male specimen of every taxon in the data matrix except for Janatella leucodesma and Ortilia liriope.

Since I found very little intraspecific variation in the samples dissected other than for the exact shape of the saccus for some taxa, I find it very unlikely that dissecting additional individuals of the taxa examined would result in changing character coding schemes for the characters coded. Fewer female specimens were dissected relative to males, and for taxa other than Chlosyne, Thessalia, and Charidryas, in most cases I dissected only one or two females per taxon.

For the genera Chlosyne, Thessalia, Charidryas, and Anemaca, all 41 Operational Taxonomic Units (OTUs) I recognize have the same character state for each character in the Melitaeini data matrix, with very few exceptions noted under the appropriate characters. Consequently, I entered these taxa into the data matrix collectively as "Chlosyne A" and "Chlosyne B". This is preferable to entering these taxa into the matrix collectively as one taxon, because a single collective OTU would not permit the analysis to show if there is evidence for the monophyly of these taxa (if two or more taxa are coded with identical states for all characters, that by itself does not imply monophyly).


Preparation of Melitaeini Genitalia for Character Analysis

I studied genitalic characters from genitalia with their natural shape intact (as

opposed to putting genitalia on slides as was done by Higgins (1960) and Higgins (1981), which involves destroying the three dimensional structure of Melitaeinine genitalia). Genitalia were dissected using standard techniques of removing a specimen's abdomen






27


and dissolving away proteinaceous material with a 10% KOH solution. The maximum magnification at which I examined genitalia was 75X, and light sources included fiber optics and/or lighting from below the microscope platform. Some preparations were stained with chlorozol back dye to highlight membranous areas.


Production of Meiitaeini Genitalia Figures to Illustrate Character States

All the genitalia illustrations (Figs. 1-292) are from camera lucida drawings. I drew these figures at 50X magnification, except for Figure 292 (drawn at 25X magnification). All the drawings were done in pencil, with differing degrees of shading used to show the difference between darker and lighter areas. I scanned each drawing, retaining its original size, with a Hewlett Packard Scan Jet ADF at 200 DPI with the output type set at grayscale. The drawings were then scanned to Adobe Photoshop 5.0.2, where I selected "Image: Adjust: Brightness/Contrast" and increased the contrast to 40%. The drawings were then saved as TIFF files, and imported into Microsoft PowerPoint, again with the original size of the drawing retained. Subsequent alterations in size (if any) as indicated on the figures, were done in PowerPoint.


Phylogenetic Analysis

Eight separate analyses were conducted with PAUP version 4.Ob4a for Macintosh, four for the cumulative out group method and four for the representative out group method. The four types of analyses included a heuristic search with equally weighted characters, a series of heuristic searches with successively weighted characters (weighting was based on character resealed consistency indices) until the tree length no longer changed, a boot strap analysis with equally weighted characters, and a boot strap






28


analysis with successively weighted characters. Strict consensus trees were calculated from heuristic searches, and 50% consensus trees were calculated from the boot strap analyses. Tree scores (consistency index, retention index, and rescaled consistency index) were calculated in PAUP based on parsimony informative characters only.

All heuristic searches were conducted with characters unordered (in my view ordering characters almost always introduces additional unwarranted assumptions into the analysis). Also, all heuristic searches were run to completion and the strict consensus trees were calculated from all of the equally parsimonious trees. For rooting options, for the cumulative out group method the "make in group monophyletic" option was selected. For the representative out group method, the option "root tree at internal node with basal polytomy" was selected.

Boot strap analyses were run with the fast stepwise addition algorithm and based on 10,000 replications. The option "retain groups with frequency >50% was used for each boot strap analysis. For analyses based on successively weighted characters, the option "sample characters with equal probability but apply weights" was selected.


Results


Characters and Character States for a Phylogenetic Analysis of the Melitaeini (Lepidoptera: Nymphalidae)

Almost all of the below character states occurring within the Melitaeini are illustrated in the camera lucida genitalia drawings presented in Figures 1-292, except for a few states particular to Phycioditi and European Melitaeiti which may be valuable to future investigations of relationships within these clades, but are irrelevant to establishing if these groups are monophyletic or how they are related to other clades. While at least one






29


figure is referenced for each illustrated character state, the below references to figures to illustrate the character states are not comprehensive; a state is often illustrated in more figures than are specifically referenced for that state. Since genitalia drawings are presented in phylogenetic order (with some exceptions to allow for more efficient use of space), and each drawing depicts multiple character states, the figure references for character states do not occur in sequence.


Characters

# =A character state which does not occur within the Melitaeiti but occurs in Nymphalini or Kallimini out group taxa.

C=For the collective out-group coding, a character state which is different from any that occurs in the Melitaeini.

Male genitalia

Characters of the valvae:

1. Hollow projection on each valve off of the inner wall, projecting off of a distinct plate on the inner posterior side of the valve (well posterior of the anterior edge of the tegumen):

#1 =Absent.

2=Present (Fig. 1).

3=Vestigial, a hollow projection is absent but a reduced plate is present in the

same position (Fig. 64).

The Kallimini generic type of Doleschalia (bisaltide) has a projection extending from the inner wall of the valvae, but it differs in position and structure from that of the Melitaeiti. I code this projection as state 3. Out group taxa Colobura dirce and Salamis augustina






30

and have hollow inner valve projections in roughly the same position on the valve as those of Melitaeini, however the shape of these projections and form of the valvae for these taxa is dissimilar to that found in Melitaeini. Some other out group taxa have dissimilar inner valve projections in different positions relative to the Melitaeinae, and many out group taxa have no inner valve projections. The collective coding for the outgroup is "1 & c", since no out-group taxa have inner valve projections that could reasonably be assumed to be homologous to those of the Melitaeini (and in any case certainly not coded as a Melitaeini state), yet for some taxa, homology can not definitively be assumed a priori to be lacking either.

2. For those taxa with state 2 of the preceding character, the general orientation of the inner valve process:

l=Posterior lateral (Fig. 1).

2=Anterior lateral (Figs. 59 & 62).

The differences between state 1 and 2 can be thought of as a hinge at the base of the valve projection, with the hinge up in state 1 and down in state 2 (contrast next character).

In Gnathotrusia mundina the inner valve process is vestigial and can not be clearly assigned to either state. I code this taxon as "?". This is the only Melitaeini examined which was found to have a vestigial inner valve process (G. epione, which I have not examined, may have this characteristic as well). Out group taxa lacking a inner valve process are coded 0, while those coded c for the preceding character are coded c for this character as well.






31

3. For those taxa with character I state 2, the curve of the posterior side of the inner valve process in ventral view:

l=Concave (Fig. 37).

2=Convex (Fig. 108).

3=Both sides fairly straight, neither distinctly concave or convex(Figs. 89 & 87). The exact angle from which the male genitalia are viewed affects the apparent shape of the inner valve process. In ventral aspect, as the anterior end of the genitalia are pushed down (increasing the posterior aspect of the view) the posterior edge begins to look concave for both state I and state 2 taxa. The state differences appear to be due to differences in how the base of the inner valve process has been rotated. Within the genus Chlosyne both states 1 and 2 occur, in addition to an intermediate state which occurs in four taxa, which does not affect this analysis. One taxon (C. leanira and its subspecies) exhibits integrades between state 2 and the intermediate state, supporting the rotation hypothesis. However, Eurodryas desfontaini appears to have independently acquired a state 2 process due to a change in shape rather than rotation, as this taxa has a forked inner valve process with the dorsal fork is in the same position as other Euphydryiti with state 1.

Since Gnathotriche sodialis and G. exclamationis have the inner valve process hinged down, the frame of reference for these taxa is different. However, if I am correct that this variation is due to the base of the process being hinged down, these taxa would be state 1. To avoid making this additional assumption, I code these taxa as "?" along with G. mundina which has a vestigial inner valve process. The out group is coded the same as for the preceding character.






32


4. For those taxa with the projection of character 1 state 2, the shape of this projection:

1 =Curved with a posterior fork (the fork in E. desfontaini is reduced to a slight

inward extending point whereas it is strong in other taxa examined with this

state) (Figs. 3 & 6-14).

2=Curved and pointed (bluntly or sharply) with a smooth simple wall (Fig. 37).

3=Dorso-ventrally flattened with five short pointed projections on the anterior

side (Fig. 43).

4=Saw-like and dorso-ventrally flattened, wedge-shaped with outer side smooth

and the inner side with approximately 8-12 prominent teeth along much of the

surface. The teeth are asymmetrical between valves and vary from three

dimensional to flattened, and single or double in the same individual (Fig. 34).

5=Approximately proximal half arched dorsally inward and roughly perpendicular

to its lateral plane, distal half abruptly changes direction, arched dorsally and projecting posteriorly inward where it tapers to a point; a short, blunt, broad

projection on the anterior side at the corner where the projection abruptly changes

direction (Fig. 81).

6=Curved and pointed and dorso ventrally flattened, edge entire except several

small but conspicuous teeth on the dorsal /outer edge.

7=B3ase broad, then narrowing, mid section serrate on sides and of similar width,

tip abruptly narrowing and then tapering to a smooth sided point.

#8=Y-shaped and covered with many prominent teeth and setae. Out-group coding is identical to the preceding character.






33


Variation of a forked inner valve projection provides a rich source of additional characters. Considerable independent evidence from other characters indicates the inner valve projection being forked is a terminal derived state. Consequently, taxa lacking a forked inner valve projection are coded "0" for the next series of characters.

5. For those taxa with state I of the preceding character, extent of the fork of the inner valve projection:

I=Strongly forked (Figs. 7-14).

2=Slightly forked (Figs. 3 & 6).

6. For those taxa with character 4 state 1, characteristic shape of the distal edge of the dorsal fork of the inner valve projection:

I1=Narrowly rounded off (Figs. 7-11 & 14).

2=Sharply pointed (Figs. 12-13).

3=Very broadly rounded off (Figs. 3 & 6).

7. For those taxa with character 4 state 1, the occurrence of prominent teeth on one side of the distal end of the dorsal fork of the inner valve process:

P=Present (Figs. 7-1 1 & 14).

2=Absent (Figs. 3,6, 12-13).

8. For those taxa with state 1 of the preceding character, the orientation of the dentate side of the dorsal fork of the inner valve process:

I Posterior (Figs. 9-1 1 & 14).

2=Anterior (Figs. 7-8).

There is no independent evidence from other characters to suggest that the presence of teeth on the dorsal fork is a terminal derived state; however, this feature is limited to taxa






34

with a forked inner valve process for which there is much independent evidence of monophyly. Consequently, Euphydryiti lacking teeth on the dorsal fork of the inner valve process are coded "?", and taxa lacking a forked inner valve process are coded "0".

9. For those taxa with character 4 state 1, the orientation of the distal part of the dorsal fork:

= Inward toward the midline (Figs. 3,6,9,11 & 14).

2=Strongly ventral, weakly inward (Figs. 7 & 8).

3=Strongly inward, weakly ventral (Figs. 12 & 13).

4=Dorsally inward (Fig. 10).

10. For those taxa with character 4 state 1, the characteristic of the base of the dorsal fork:

l=Narrow (of similar width to the tip), elongate and not or very slightly tapering

(Figs. 7-8).

2=Broad (distinctly wider than the tip), short, and tapering (Figs. 3,6,10-12 & 14).

3=Narrow, tapering, and elongate (Fig. 13).

The orientation of the tip of the ventral fork falls into discrete states for those taxa with a forked inner valve projection. However, some of the taxa with a forked projection such as E. desfontaini and E. aurinia suggest the dorsal fork could be thought of as a process off of the ventral fork, and that the ventral fork is likely homologous to the inner valve process of other Melitaeini while the dorsal fork is derived. Consequently, it would be invalid to assign derived states of the ventral fork to the Euphydryiti group and code remaining taxa 0. Orientation of the tip of the inner valve process does not fall nicely into discrete states throughout the Melitaeini, however, this is no reason to discard this






35


information within the Euphydryiti. For the following character, I code taxa lacking a forked inner valve process as "?". 11. For those taxa with a forked inner valve process, the orientation of the tip of the ventral fork:

l=Curved ventrally outward (Figs. 7-8).

2=Curved ventrally inward (Figs. 3 & 9-14).

3=Strongly inward and slightly ventrally (Fig. 6).

The variation within taxa with a saw-like toothed inner valve process (character 4 state 4) can be further subdivided into discrete states. Independent character evidence suggests that this form of the inner valve process is a terminal derived state. 12. For those taxa with a saw-like inner valve process, the distal extent of the teeth on the process:

1=Distal half free of teeth.

2=Teeth continue distinctly distal of the distal half (Fig. 34). Taxa lacking a saw-like inner valve process are coded 0. 13. For those taxa with a inner valve process (character 1 state 2), the shape of distal end of this process:

I =Narrow and sharply or bluntly pointed (Fig. 1).

2=Flared out and serrate (Figs. 10-14).

Most state I taxa examined have a sharply pointed process; however, it is bluntly pointed in some derived Chlosyne. This variation is addressed in the data matrix for the Chlosyniti.





36

14. A distinct hollow ventral valve process at the posterior end of each valve, in part anterior to the part of the valve with extensive long hair-like setae:

0=Absent (Figs. 15,39,45,60,64.1-64.2 & 76).

1=Present (Figs. 1,36,42?,82,85,91,96,101,114,116,118 & 120).

Coding this character for some European taxa is problematic. The Cinclidia examined have a broad posterior part of the valve extended as a posterior process. At the base of this process, there is a ventral hollow projection which appears to be slightly posterior to the position of the other state 1 taxa. The most objective coding in my judgement is to code this taxon as state 1. However, for characters pertaining to the form of the ventral valve process, I code the taxa in Cinclidia as "?" to avoid weighting a potentially independently acquired character state twice. While the Mellicta examined all have a ventral process in the same position as the other state I taxa, this problem does not apply as the form of the process in Mellicta is different from all other state I taxa. The process at the posterior ventral comer of the valve in D. didyma and D. trivia is formed from the part of the valve with hair like setae, as opposed to a process anterior to the part of the valve with hair like setae, and is therefor coded as a state of character 18.

Most of the out group taxa lack any ventral valve process. However, a few taxa have a ventral valve process which is quite dissimilar to anything found in the Melitaeini in terms of structure and position. I code the out-group taxa collectively as "O&C." 15. For those taxa with a ventral valve process, the curve of the ventral valve process:

I=A posteriorly curved pointed process with a convex anterior side and concave

posterior side (Figs. 1,82,85,91,96,101,114,116,118 & 120).






37


2=A process of variable shape that's distal most point curves anteriorly

(sometimes only slightly) with the anterior wall of the distal most point concave

and the posterior wall convex (Fig. 36). Taxa lacking a ventral valve process are coded 0. 16. For those taxa with a ventral valve projection of character 14, the variation in shape of the process among state I and state 2 taxa for the preceding character can further be divided into discrete states:

I =Curved and pointed with the walls entire (or with one tooth-see the following

character) (Figs. 1,85,91,96,101,114,116,118 & 120).

2=Anterior edge smooth with 1-2 small teeth distally, posterior edge serrate with

several (approximately 6) distinct teeth (Fig. 82).

3=Veryjagged, anterior portion of process with 2-3 prominent teeth in a row

(some additional smaller teeth may be present in-between) with a similar distal

extent, posterior portion of process extends much farther distally as a curved,

pointed projection with 0-1 smaller teeth on its anterior edge and 1-2 smaller teeth

on its posterior edge (posterior to the longest prong there may be 3 or 0 short

teeth) (Fig. 36).

4=A deeply forked projection, with the longest (anterior) fork curving anteriorly

and the shorter (posterior) fork curving posteriorly.

5=A single broad process with a serrate anterior side and an entire posterior side. States I and 2 are for taxa which have state I for the preceding character, while states 3-6 are for taxa which have state 2 for the preceding character. The above coding avoids biasing the analysis for or against the homology between these two states of character 14,






38


and avoids the unnecessary question mark coding that would result from splitting this into two separate characters. Taxa with no ventral valve process are coded 0. Within the group of taxa with character state I for the preceding character, the variation can be further subdivided into discrete states. 17. For those taxa with character 6 state 1, the presence of a small additional distal tooth at the distal end of the process:

I =None.

2=A small tooth is present just proximal to the distal end on the anterior side

(Fig. 107).

3=A small tooth is present just proximal to the distal end on the posterior side

(Fig. 94).

Taxa with a ventral process but lacking state I for the preceding character are coded "?". Taxa lacking a ventral valve process are coded 0. 18. Posterior part of valve (area with many hair-like setae) extended to form a distinct pointed valve projection, located posterior to the ventral projection of character 5 (if present).

I =None (See also note for character 3 1

2=Posterior ventral side extends as an elongate, narrow, hollow process (Figs.

1,85,115,117,119 & 121).

3=A flattened non-hollow projection is present (Fig. 93).

4=Posterior dorsal side extended as an elongate, narrow, hollow process (Fig. 39).






39


5=Posterior side of valve with multiple hollow processes, with the most

prominent one at the posterior ventral corner, and the next most prominent at the

posterior dorsal corner (Fig. 45).

Some Nymphalini and Kallimini with valvae very uncharacteristic of Melitaeini have posterior valvae projections which appear highly dissimilar to any of those found in Melitaeini. Such taxa are coded as C for this character. 19. A broad area of the posterior end of the valvae, including the area with dense hairlike setae and the barer area ventral to it, expanded into a broad posterior valve projection:

0=None.

I=A broad dorsally curved projection with a convex ventral surface and a concave

dorsal surface (much shorter than the dorsal surface), bearing two prominent

ventral projections (Fig. 42).

20. Inner wall of ventral part of valvae turned roughly 90' and lobed, forming a pocket shaped roughly like a half circle with many spine-like setae on the posterior surface.

I =Absent.

2=Present (Figs. 2,3,5 & 15).

The shape and lobing of this structure appears somewhat different in Eurodryas desfontaini from the other taxa with this structure; however, a structure with identical spine-like setae is present on the same part of the valve.

I considered coding this variation as two characters, one for the spine-like setae and another for the lobe of the valvae; however, since the spines occur on the lobe and all






40


taxa with the lobe have spines, there would seem a good chance that these characters are not independent.

2 1. For those taxa with state 2 of the preceding character, the spine like setae are:

I =Numerous, dozens are present (Figs. 2,3 & 15).

2=Only five, especially large ones are present (Fig. 6).

Independent evidence from other characters suggests state 2 of the preceding character is a terminal derived state; therefore taxa lacking this state for the preceding character are coded 0.

22. The shape of the area containing dense setae on the inner posterior side of the valve:

Il=A roughly bean-shaped plateau with a flat surface (except for setae sockets).

2=The area is curved and not forming a plateau.

3=The area forms a plateau with a broad ventral side, widening slightly and then

tapering to almost a point on the dorsal side.

4=The area forms a plateau with a narrow, rounded ventral side and a broadly

rounded dorsal side.

All Melitaeini have an opening between where the plates of a valve fold around and overlap, always visible in ventral view. In most taxa, this opening ranges from not visible to barely visible along the ventral edge in lateral view. However, the position of the opening is notably different in Euphydryiti. 23. Position of the above ventral valve opening in lateral viewv:

I =Completely and clearly visible in lateral view dorsal to the ventral edge of the

valve (Fig. 15).






41


2=Partly visible to not visible in lateral view at the ventral edge of the valve (Figs.

42,45,114,116,120).

24. The form of the above ventral valve opening:

l=A distinct opening (Figs. 1,2,34,37,40,43,58,61,63,75,80,94,108,110,112 &

122).

2=A slit (Figs. 87,89 & 106).

25. The anterior extent of the above opening with respect to the ventral posterior edge of the vinculum:

I1=Terminates well posterior to the vinculumn (Fig. I & all other figures of ventral

male genitalia capsules).

2=Terminates anterior of the vinculumn (Fig. 80).

There is considerable variation in the anterior extent and length of the above valve opening and in the posterior extent of the vinculumn within the Melitaeini, most of which does not fall into discrete states. However, only taxa in the genus Atlantea have the vinculumn overlapping the anterior ends of the valve, including the ventral valve opening.

All Melitaeini have a patch of setae on the inner side of the valve (around the

vicinity of the anterior to posterior midpoint). The extent of folding of the valvae in this area determines whether the sockets for these setae are visible in ventral view. 26. The visibility of the sockets for the above setae in ventral view.

1 =Visible (Fig. 1 & all other figures of ventral male genitalia capsules).

2=Concealed (Figs. 99 & 106).

Note that some Phycioditi have the setae sockets just visible at the inner edges of the valve in ventral view. The out group is coded "?" for this character, because there is a






42


great diversity of valve structure and setae arrangements within the Nymphalini and Kallamini not found within the Melitaeini. 27. Posterior ventral inner wall of each valve extended inward in a flat, triangular projection with the anterior side approximately perpendicular to the midline of genitalia and the posterior side slanted inward.

1 =Absent.

2=Present (Figs 58 & 62-63).

28. Proximity of the valvae to each other in ventral and dorsal view:

0=Valvae distinctly separated (Fig. 1 and all other figures of male genitalia

capsules in ventral or dorsal aspect).

I =Valvae touching each other or nearly so (Figs. 34-35).

The only taxon that was somewhat ambiguous in scoring this character is Mellicta aurelia. In this taxon the valvae are still distinctly separated, but not by as much as other state 0 taxa.

29. The presence of a patch of short setae on the dorsal half of the posterior lateral sides of the valvae (distinct from the long setae at the posterior edge of the valvae):

0=Present (Figs. 15,36,39,42,45,64.2,76,82,96,101,114, & 176-192 (minus

178,179,183 7 187).

l=Absent (Figs. 60,64.1,116,118 & 120).

This character can be difficult to detect with a camera lucida setup, and it was inadvertently omitted from some of the first figures I produced of lateral genitalia capsules, including Figs. 91,178,179,183 & 187. These taxa actually do have the patch of






43


short setae characteristic of state 0, although these setae are not illustrated by these figures.

30. Posterior tapering of the entire valvae (This does not refer to a distinct posterior projection off of an otherwise non-tapering valve):

I1=Valvae do not extend and taper to a point posteriorly (All Figs. of lateral

genitalia capsules minus those indicated below)..

2=Valvae extend posteriorly and taper to a point, with the area of hair-like setae

located along this extension (Figs. 60,64.1,64.2 & 76).

#3=Valvae distinctly tapering but nowhere near to a point as in state 1, ends of

valvae broadly rounded.

3 1. For those taxa with state 2 of the preceding character, the characteristic of the posterior terminus of the valvae:

I1=Posterior end of valve with a short, curved, pointed tooth (Fig. 75).

2=Posterior end of valve with a long, narrow, tubular extension with a flared,

flatter end (extension arches dorsally, tip points ventrally) (Higgins 198 1, Fig. 275

pg. 207).

3=Long, tapering tubular extensions ending in a sharp, inward curved point (Figs.

58-64.2).

Taxa lacking state 2 of the preceding character do not have a distinct posterior valve terminus. Evidence suggests state 2 of the preceding character is a terminal derived state; however, taxa lacking this state are coded "'?". This is because some taxa without a valve that tapers to a point have a distinct posterior projection off of a non tapering valve (character 18). An alternative coding scheme could be to code the above three states as






44


three new states for character 18. An analysis with this alternative coding scheme did not affect the resulting consensus tree generated by PAUP. 32. In lateral view, the orientation of the valvae:

0=Much more posterior than ventral (Fig. 76).

1 =Much more ventral than posterior (All other figures of male genitalia capsules

in lateral aspect).

32.1. Outer lateral side of each valve with a groove, originating near the anterior-dorsal comer of the valve and slanting posterior ventrally across the valve for much of its length.

0=Absent.

I =Present (Figs. 60 & 64.1). Characters of the juxta:

A number of Melitaeini have various ridges or plateaus on the ventral surface of the juxta, while others have a predominantly smooth featureless juxta. The nymphalid out groups have the juxta features quite different from the Melitaeini. 33. Ventral surface of the juxta:

I =With a mid ventral ridge extending to the posterior edge of the juxta and

originating a variable distance anteriorly (Fig. 5).

2=Fairly smooth and lacking distinct ridges or plateaus (Figs.

4,34,40,43,61,63,75,80,87,89 & 94).

3=Narrow ridge tapering anterior to posterior along the midline, flared anteriorly

into a broad triangular plateau (Fig. 37).






45


4=A prominent raised plateau composed of a diamond-shaped posterior section,

and widening anteriorly into a broad triangular section with slightly concave

sides. The lateral sides of the plateau are steep and distinct while the anterior and

posterior sides more gradually slope dorsally to the level of the remainder of the

juxta (Figs. 1, 108 & 134-155) .

5=A distinct posteriorly directed triangular-shaped plateau, including the entire

anterior part of the juxta as the base of the triangle; the sides and tip are bordered

by steep dorsal slopes (Figs. 1 10 & 122).

6=A distinct raised plateau with the anterior side narrow and transversing the

width of the juxta, the middle greatly constricted with deeply concave sides

adjacent to steep dorsal slopes, and the posterior section broad and gently sloping

dorsally (Figs. 112-113).

7=Roof-shaped, the ventro-anterior face is triangular (pointed ventrally) with a distinct boundary, and each side of the posterior part is rectangular and slanted

ventrally inward, with the two sides meeting along the ventral midline (Fig. 99).

8=Smoothly curved anteriorly, roof-shaped with a sharp ventral keel in

approximately the posterior 4 (Fig. 94).

One of the taxa with state 5 has the apex of the plateau extended as a posteriorly directed spine. This variation is not coded, as only one taxon has this feature. The spine is dissimilar to that found in most Euphydryiti, which possess a posterior juxta projection composed of two parts close together.

34. For those taxa with state 1 of the preceding character, the anterior extent of the mid ventral juxta ridge:






46

l=Terminating distinctly before the anterior edge of the juxta (Fig. 5).

2=Extending to the anterior edge of the juxta.

All these taxa have the ridge keeled posteriorly, but those with the ridge extending anteriorly to the anterior margin have this part of the ridge smooth. This variation coincides exactly with the above character, thus it is not coded since it does not appear to constitute independent variation. Independent evidence from other characters suggests that state 1 of the preceding character is a terminal derived state, so taxa lacking this state are coded "0" for the above character. 35. The presence of dense, minute, granulose structures on the ventral midline and other portions of the ventral surface of the j uxta:

0=Absent.

l=Present (Fig. 5).

36. Ventral posterior midline ofjuxta terminating in two distinct, ventral-posteriorly projecting points (an anterior and posterior component). The two projections are very close together and appear to be a single projection unless viewed from the lateral side, where the separation can be detected.

I =Absent.

2=Present (Figs. 2 & 5).

37. For those taxa with state 2 of the preceding character, the relative sclerotization of the posterior projections of the juxta:

l=Far more heavily sclerotized than the remainder of the juxta, appearing almost

black.






47

2=Somewhat more heavily sclerotized than the remainder of the juxta, however

not nearly to the degree as state I and clearly not appearing almost black (Figs. 2

&5).

Since this character is scored by the sclerotization of the juxta relative to that of the projection in the same specimen, differences in how long different specimens were left in KOH should not prevent the ability to accurately score this character. However, the specimens scored for this character were placed in KOH from the same stock solution at the same time, and left in KOH at the same temperature (room temperature) and then dissected in succession (except for Hypodryas iduna). However, I found other dissected specimens left in KOH for differing amounts of time could still be coded the same way without ambiguity.

Taxa lacking character 18 state 2 are coded 0, with the exception of Eurodryas desfontaini. Many other independent characters support the monophyly of the Euphydryiti clade, and E. desfontaini is the only Euphydryiti examined which lacks the derived state of the preceding character. I avoid coding this taxon as 0 for the above character, because the absence of paired juxta projections is the only character suggesting E. desfontaini may be more primitive than the other Euphydryiti examined. Coding E. desfontaini as 0 rather than ? would produce an unwarranted bias. 38. For those taxa with paired posterior juxta projections, the lateral compression of these projections:

l=Distinctly laterally compressed (Fig. 5).

2=Not laterally compressed (Fig. 2).






48

Taxa lacking paired posterior juxta projections are coded the same as for the preceding character.

Characters of the saccus:

39. Forking of the saccus due to invaginated projections:

O=Unforked.

l=Forked (All figures of male genitalia capsules in dorsal and ventral aspects). 40. Further characterization of anterior-dorsal projections of the saccus formed from invaginations of the saccus:

1 =Two bilaterally symmetrical prominent projections, one on each side (Figs.

1,2,5,34,37,40,44,59,61,63,74,81,94,99,106,108,111,134-173).

2=Saccus tapering anteriorly and triangular, then forming a narrow extension

forked with two small extensions at the anterior most end (Figs. 86 & 89).

3=A single prominent projection to the right of the midline in ventral view

(Higgins 1981, Fig. 302 pg. 211).

4=Saccus extends in a centered rectangular projection, slightly forked at the

anterior end (Higgins 1981, Fig. 272 pg. 207).

5=A single prominent centered projection (Higgins 1981, Fig. 416 pg. 227).

6=Similar to 5, except the central projection is slightly forked at the anterior end

(Higgins 1981, Fig. 475 pg. 236).

7=Saccus roughly triangular with an abrupt constriction at about one-half its

length, slightly forked at anterior most end (Fig. 112).

#8=Entire saccus projects anteriorly in an elongate, hollow, extension, as opposed

to any distinct invaginated projections from the saccus as in states 1-7.






49


#9=Entire saccus projects anteriorly in a broad, triangular extension.

#A=Saccus broad and emarginate anteriorly with no extensions.

These character states are subdivisions of state 0 and state I of the preceding character. The above approach as opposed to splitting this variation up into two characters (one for state 0 taxa and one for state 1 taxa of the preceding character) avoids unnecessary "?" coding.

There is great variation among the taxa with paired saccular projections (state 1), but this variation was not found to fall into discrete states. In addition, exact shape of the saccus was found to exhibit considerable intraspecific variation within the Chlosyniti (Figs. 167&169, & 225-260). The forms of character states 2 and 7 were found to be consistent, although there is some interspecific variation within state 2. States 3-6 are particular to Phycioditi, and saccus shapes of Phycioditi are illustrated in Higgins (1981). Also, note some Phyciodes have a single centered saccular projection as do some Nymphalid out group genera. While all are coded state 5, the shape and length of these projections between the Phycioditi state 5 taxa and out group state 5 taxa varies greatly. 41. For taxa with a pair of saccular projections, the presence of a broad, flat plate of lightly sclerotized tissue between the projections:

1 =Absent.

2=Present (Figs. 58-59,61-62 & 63-64.1).

Overwhelming evidence from other characters indicates a forked saccus is a derived state of a binary character for the Melitaeini and those Melitaeini with a non-forked saccus underwent a reversal. Since the above character can not be coded for taxa lacking a






50

forked saccus, I code the Melitaeini without a forked saccus as "?" and the out group as

0.

42. Given state 2 of the preceding character, the points of attachment of the connecting plate of tissue between the saccular fork:

1 =Anterior margin connects across the distal ends of the fork (Figs 64 & 64.1).

2=Anterior margin connects across the vicinity of the midpoint of the fork (Figs.

59 & 61).

Independent evidence from other characters suggests state 2 of the preceding character is a terminal derived state. Taxa lacking this state are coded "0" for the above character. 43. Extent of the saccus at the middle of the ventral anterior side of the genitalia:

I =Prominent/well developed.

2=Reduced to a very narrow bridge (Fig. 2).

44. Exposure of openings to the hollow projection(s) of the saccus:

1 =Openings concealed by vinculum (Figs. 1,5,34,37,44,43,58,61,63,80,

87,89,94,99,108,110,134-155).

2=Vinculum anterior of openings leaving them exposed (Figs. 75,112 & 122).

#3=Vinculum slightly posterior or even with saccular projections posterior opening, however the vinculum is orientated so far ventral that the saccular

projection's openings appear to be exposed. Characters of the tegumen/uncus:

45. Development of the uncus:

O=Tegumen expanded into a well developed uncus (Figs. 35,59,62,74,86,90,95,

99,106,109,111,113,123,156-173).






51


1=Tegumen reduced to a thin, simple, sclerotized bridge between the valvae

(Figs. 1,4,16-23,25,38,41.

Within the state 0 taxa (nymphalid out groups and many Melitaeini) there is considerable variation in the shape and size of the well developed uncus, which I was not able to reliably code into discrete states. There is relatively little variation in the simple tegumen of state I taxa, other than for its exact width. While Euphydryiti have prominent posterior projections on the tegumen, it is otherwise a simple bridge characteristic of state

1.

There are some taxa which were somewhat ambiguous in scoring this character, in particular Gnathotriche mundina and Atlantea. I code these taxa as "?". Higgins (1981) illustrates a well developed conical tegumen for G. mundina, however, the sole specimen of this taxon I was able to examine has the posterior border of the tegumen straight across (Fig. 64) with the juxta shape more reminiscent of Higgins (1981) illustration of the tegumen. Since I have found tegumen shape to exhibit very little intraspecific variation within the Melitaeini (including in G. exclamationis), and since Higgins greatly compressed the genitalia he examined by placing them on slides, I hypthesize that the tegumen illustration for G. mundina in Higgins (1981) is in error. 46. Each outer side of tegumen with a very heavily sclerotized plate containing a raised, pointed lateral ridge:

1 =Absent.

2=Present (Figs. 100-102).

47. Clusters of small, heavily sclerotized, pointed teeth covering each posterior corner of the uncus:






52


1 =Absent.

2=Present (Higgins 1981, Fig. 475 pg. 236).

48. A heavily sclerotized convexly curved claw on each posterior side of the uncus:

1 =None.

2=An entire claw is present (Figs. 74 & 76).

3=A dentate claw is present (Higgins 1981, Fig. 473 pg. 235).

An alternative coding scheme would be to code for presence or absence of a claw, and then have a separate character for whether the claw is smooth or dentate. However, the structure of the dentate and entire claws and their exact position is quite different, so the above coding seemed most appropriate. It also may be appropriate to combine the two preceding characters as a single multistate character. These character states are particular to the Phycioditi, and not informative other than for relationships within the Phycioditi, and consequently extraneous to the goals of this analysis. Examination of more Phycioditi taxa would give more insight into the range of variation occurring within Phycioditi and the best way to code it.

49. Hollow, paired, pointed posterior projections formed from invaginations of the inner wall of the tegumen (some of the out-group states refer to single projections-only paired projections occur in the Melitaeini):

O=None.

1 =Projections divide the tegumen into a deep fork; they are connected by only a

very thin bridge of tegumen between them and anteriorly (Figs. 4,16-23 & 25).






53


2=Projections do not divide the tegumen, but extend from its sides and are connected by a longer and much wider (relative to state 1) area of tegumen on their anterior side (Figs 35,80,81,86,90 & 95). 3=Ventral lateral inner side of tegumen extended as long, narrow, hollow invaginations covered extensively with many small teeth on the inner surface (Fig. 64).

4=Ventro-lateral outer sides of tegumen extended as minute triangular pointed projections (Fig. 44).

#5=A large, deeply forked posterior projection arising from the center of the tegumen.

#6=Paired prominent projections extend from the middle of the tegumen (versus the sides in state 2), one on each side of the dorsal midline with very little space between them, a wide area of tegumen occurs anterior to these projections. #7=Entire tegumen extends posteriorly in an elongate pointed posterior projection.

#8=Entire posterior tegumen deeply forked, occurring as two elongate tapering pointed projections.

#9=Entire posterior tegumen extends as an elongate (concave in posterior view) spatulate projection.

#A=A single elongate tapering posterior pointed projection at the midline. #B=Entire tegumen extends posteriorly as an elongate tapering projection but with a broadly rounded terminus.






54


#D=A single elongate centered projection widened posteriorly with two lateral

pointed extensions.

Due to the notably different structure/characteristics and position between the invaginations of the above in group states, I did not code for the presence or absence of paired invaginated projections of the tegumen. 50. For those taxa with character 49 state 1, the shape of the tegumen projections:

l=Elongate, narrow, and curving ventrally to a point distally (Figs. 23-24).

2=Short, broad, and triangular with the outer side slanted and the inner side nearly

vertical (Fig. 21).

3=Prominent and triangular with a broad base and a narrow straight tip (Fig. 16).

4=Base broadly triangular but the distal portion narrower and slanting inward

toward the midline; the distal portion is blunt and more finger-like than triangular

(Fig. 19).

5=Long and pointed with the outer side concave and the inner side curving

outward distally while being more vertical/concave basally (Figs. 18,20,22 & 25).

6=Prominent, proximal approximately 2 broad and triangular, distal 2 narrower

and triangular curving sharply outward at the distal extremity (Fig. 4).

7=Short and broad, basal 2 somewhat rectangular, distal 2 projects from lateral

side of base, triangular, and with the inner side concave and outer side fairly

straight (Fig. 17).

Independent evidence from other characters suggests state 1 of the preceding character is a terminal derived state. Consequently, taxa lacking this state are coded "0".






55


5 1. For those taxa with character 49 state 1, the presence of an additional pointed projection below the apex on the inner sides of the tegumen projections:

I =Ab sent.

2=Present (Figs. 19 & 2 1).

52. For those taxa with character 49 state 2, further characterization of the projections off of the tegumen:

I=About as long as length of tegumen anterior to the projections or slightly

shorter, for most of length at most slightly tapering; however, apically projections

curve inward a short distance and taper abruptly to a point (Fig. 35).

2=About as long as length of tegumen anterior to the projections or slightly

shorter, outer side convex and inner side straighter and slightly concave, apex

pointed (Fig. 2).

3=Small, triangular, and much shorter than the length of the tegumen anterior to

the projections, apex pointed (Figs. 86 & 90).

4=Weakly triangular and much shorter than the length of the tegumen anterior to

the projections, outer side somewhat convex, inner side slightly convex

proximally and straight distally, apex blunt (Fig. 80).

Independent evidence from other characters suggests state 2 of the third preceding character independently arose twice; however, there is no overlap of the above states between the two clades where this state arose. Consequently, taxa lacking state 2 of the third preceding character can be coded 0 without risk of doubly weighting an independent acquisition.






56


52.1. For those taxa with character 49 state 2, the presence of setae on the tegumen projections:

I =Present.

2=Absent.

As noted, independent evidence from other characters suggests character 49 state 2 independently arose as a terminal derived state in two lineages, and taxa in each lineage have the state of setae absent. To code them both as such has the danger of weighting a likely independent acquisition twice. However, this character is nonetheless informative within the lineage that includes taxa with and without setae on the projections (Poladryas, Atlantea and Higginsius), and the character should not be discarded. I code these taxa for the above character, those Mellicta with character 49 state 2 and taxa with other tegumen projections as "?", and other taxa as "0". All Euphydryiti have minute granulose patches on their tegumen projections. These are the only taxa with these granulose patches on the tegumen, and there seems no reason to assume the presence of these patches is dependent on having a particular type of tegumen projection. In some taxa these patches are just detectable at 50X magnification with good lighting, while in others they are conspicuous, with some apical granulose patches enlarged into teeth-like patches.

53. The presence of minute granulose patches on the tegumen:

0=Absent.

I =Present.

54. For those taxa with state I for the preceding character, the apparency of these patches at 50X magnification with good lighting:






57

l=Very conspicuous (Figs. 19, 23 & 24).

2=Obscure (generally not detectable with the camera lucida on).

Independent evidence from other characters suggests state 1 of the preceding character is a terminal derived state, and taxa lacking this state are coded "40". Characters of the Phallus:

55. Laterally flattened keel on the anterior and anterior-ventral end of the phallus:

1=Present (Figs. 28,48-50,57,66-67,72,78,84,88,92,97-98,103-105,126-133,213253).

#2=Absent.

#3=Absent, but a flared wedge-shaped hollow structure is present.

#4=Absent, but a hollow, open, ventral scoop-shaped structure is present.

56. Orientation of the posterior opening of the phallus relative to the anterior opening:

1 =Posterior opening is ventral and approximately 180 degrees around from dorsal

anterior opening (Figs.26-29 and all other phallus figures).

2=Anterior opening is lateral and approximately 90 degrees around to the left (in

ventral view) with respect to the ventral dorsal opening (Figs. 30-33).

#3=Anterior opening is neither distinctly dorsal or ventral, posterior opening is

ventral.

#4=Posterior and anterior openings are both dorsal.

All Melitaeini have a pattern formed by areas of lightly and more heavily sclerotized tissue on the ventral posterior surface of the phallus, posterior to the supersensory membrane. Variation can be coded with respect to the pattern of the lightly sclerotized area, and whether it extends anteriorly beyond the supersensory membrane.






58

57. Sclerotization pattern of the ventral side of the phallus posterior to the supersensory membrane:

I =Elaborate spear-shaped design (Figs. 1 & 213-233).

2=Posteriorly lightly sclerotized all the way across, anterior to this area the darkly

sclerotized area begins along the sides, widening until extending all the way

across forming a concave border (Figs. 26,30,49-51,53 & 104).

3=A narrow band of dark sclerotization on each side, and lightly sclerotized in the

middle (Figs. 65,68,73,79,88,92 & 98-99).

4=Lightly sclerotized area ovoid, with darkly sclerotized area reduced to an

extremely narrow band on each side at one point, and expanding anterior and

posterior to this point (Figs. 127,129,131,133).

5=Similar to state 2 except for a narrow, posteriorly tapering triangular section of

more heavily sclerotized tissue originating at the supersensory membrane, and

the border of between the lightly and darkly sclerotized areas is convex (Fig. 28).

6=Like state 3 except narrow inward anterior slanting extensions of sclerotized tissue from the sides (just anterior to the plate attached to the vesica) (Fig. 97).

Several taxa, Euphydryas phaeton, E. editha, and the E. anicia complex have the phallus twisted 90 degrees such that the lateral side is equivalent to the ventral side for other taxa. This information is coded as a separate character, and assigning a new derived state for the ventral sclerotization pattern would weight the same feature twice. The lateral side of these taxa is typical of state 2, and they are coded as such. The same applies to character 60 below.






59

58. For those taxa with state 2 for the preceding character, the point of the posterior extent of the more heavily sclerotized area:

I =Much closer to the posterior opening than the supersensory membrane (Figs.

49-51 & 53).

2=Closer to the supersensory membrane than the posterior opening (Figs. 26,30 &

104).

Taxa lacking state 2 for the preceding character are coded "?" 59. The anterior extent of the lightly sclerotized area on the ventral phallus surface:

0=Reaching (and may surpass) the supersensory membrane (Figs. Figs. 26,30,4951,53,65,68,73,79,88,92,97-98,104,127,213-214,216-224).

I =Terminating distinctly posterior to the supersensory membrane, but closer to the supersensory membrane than the posterior phallus opening (Figs. 129,131,

133, 225-227,230,232-233).

2=Terminating distinctly posterior to the supersensory membrane, but closer to

the posterior phallus opening than the supersensory membrane (Fig. 53).

This character varies within the genus Chlosyne. I code Chlosyne as "?" in this analysis but code Chlosyne taxa individually for this character in the Chlosyniti data matrix. 60. Sclerotization pattern of the dorsal surface of the phallus:

l=The longitudinal middle is more heavily sclerotized relative to the sides, and

this sclerotized area tapers anterior to posterior (within this area the midline may

appear lighter than the sides) (Figs.29 & 31).

2=Uniformly sclerotized (the sides may appear darker in dorsal view due to their

proximity to sclerotized area of the phallus on the ventral side) (Figs. 69 & 77).






60


61. A hollow tube open on both ends on the distal posterior end of the phallus.

I =Absent.

2=Present (Figs. 83,88,92,97 & 98).

62. Posterior dorsal end of phallus distal to the posterior ventral side: I =Entire scierotized area on posterior dorsal side of phallus in the form of an elongate

triangular tapering area extending distinctly distal to the ventral side of the phallus

opening (Figs. 29-30).

2=Not extending distinctly distal to the posterior ventral end (Figs. 49-5 7 & 77-79).

3=With an elongate tapering extension with the sides slightly concave (Figs 88,92 & 9798).

4=A triangular extension with a squared off tip (Figs. 65 & 7 1).

5=A broad extension with a convex base (Figs. 104,127,129,131,133 & 213-233).

63. For those taxa with state 5 of the preceding character, the presence of a triangular projection at the distal posterior edge of the phallus extension:

I1=Present (Fig. 104).

2=Absent (Figs. 127,129,131,133 & 213-233).

Independent evidence from other characters suggests state 5 of the preceding character is a terminal derived state with one independent acquisition in G. exciarnationis, therefor taxa lacking this state are coded "O"and G. exclamnationis is coded "?". 64. For those taxa with an extension of the dorsal posterior side of the phallus distal to the ventral posterior side, the sclerotization pattern on the ventral side of this extension:

I1=Sides and middle darkest (the sides may be as dark or darker than the middle) with the sides and middle separated by an area of lightly sclerotized tissue of variable width (best

illustrated in Figs. 1 & 228, difficult to see in some figures of state I taxa because the






61

lightly sclerotized area between the sides and middle is narrow and requires high

magnification to illustrate).

2=A triangle within a triangle pattern, with the arrow head pattern in-between lightly

sclerotized (Fig. 104).

3=Uniformly sclerotized except may appear darker on the sides where they are folded

over (Figs. 26,28,30,65,69,71,88,92 & 97-98).

Taxa with no extension on the dorsal posterior side of the phallus are coded "?". The following characteristics associated with the vesica were examined by everting the vesica for specimens of all included species with the exception of some Phycioditi. Within the Phycioditi, which are notoriously difficult to evert the vesica, I found the character states of these characters to be invariant, and for those taxa where I did not evert the vesica I still was able to score the state for every character except for the presence of teeth on the ventral surface of the everted vesica. I was also unable to evert the vesica on some of the out-group taxa. Failure to evert some of the out group taxa was problematic for some of the following characters but not others, as for some characters it was clear out group taxa do not contain the pertinent structures (such as plates on the posterior end of the phallus attached to the vesica) and consequently lack any forms of these structures. For the analysis where Anartia, Junonia, and Catacroptera were used as out groups, the above complications do not apply because the vesicas of all these taxa were everted.

65. The presence of minute teeth on the ventral surface of the everted vesica:

I=Present (Figs. 52 & 54-57).

2=Absent.






62


State 1 is not to be confused with a sculpturing of small bumps (as opposed to distinct pointed teeth) which occur in the Euphydryiti. None of the out group taxa examined have teeth on the vesica; however, since some outgroup taxa were not everted the out group is coded "?" for this character.

The presence of the above minute teeth appears to be a unique synapomorphy for European Melitaeiti; however, the teeth may be absent in D. trivia. I was unable to confidently assign a state to D. trivia even under maximum power-this taxon may have extremely minute teeth present but I could not be sure. If D. trivia does have these teeth, they are greatly reduced relative to any of the other state I taxa examined. At present, I code D. trivia as "?". Other taxa where this character is coded "?" are taxa for which I did not evert the vesica, but in all likelihood would have lacked teeth on the ventral everted vesica.

66. Proximal part of the everted vesica encircled by dense, minute, granulose patches (conspicuous at 50X magnification with bright lighting).

O=Absent.

l=Present (Figs. 27-28 & 33).

The out group is coded "?" for the same reasons as for the preceding character, although none of the vesicas examined for out group taxa were found to possess this feature. 67. Given state I for the preceding character, the length of the band of granulose patches encircling the vesica relative to its width:

1 =Distinctly wider than long (Fig. 28).

2=Distinctly longer than wide (Figs. 27 & 33).






63


Taxa lacking state I for the preceding character are coded "?", since the failure to evert the vesica on some out group taxa precludes reliably concluding that the presence of the granulose patches is a terminal derived state. 68. The presence of a pair of disjunct sclerotized patches containing minute teeth on the ventral surface of the everted vesica:

O=Absent.

1=Present (Fig. 28).

For the two taxa examined which possess state 1, in E. desfontaini the sclerotized patches are shorter and more distal, while in E. aurinia they are longer and more proximal. No out group taxa were found to possess state 1; however, since some out group taxa did not have the vesica everted the out group is coded "?". 69. Sclerotization of ventral part of vesica, viewed when everted:

1=None, this area is clear and membranous, staining blue if chlorozol black is

applied.

2=Present, this area does not stain blue.

#3=Many thin, longitudinal, lightly sclerotized bands proximally on vesica for a

length exceeding /2 the length of the phallus.

For those taxa designated as not having been examined with the vesica everted (also not stained), this character was examined at high power by placing the phallus in a glass petri dish with a bright light source underneath. With this technique, I found I was able to score state 1 and state 2 specimens prior to everting the vesica. 70. A sclerotized plate on the vesica (best seen when vesica everted) attached to each ventral lateral side of the phallus.






64


1=Absent (Figs. 27-28 & 33).

2=Present (Figs. 48,54-56,66,69,78,83-84,105,112 & 211-212).

71. For those taxa with state 2 of the preceding character, a second plate, bearing teeth, attached to the vesica and connected to the distal end or distal-ventral comer of the plate attached to the phallus :

l=Absent (Figs. 27-28,33,54 & 66).

2=Present (Figs. 48,52,55-56,69,78,83-84,105,112 & 211-212).

The structure of the plates attached to the vesica in Gnathotriche sodialis is very unusual. A single teeth-bearing plate is attached to the phallus over a broad area, including the entire posterior sclerotized end of the phallus from the point where the phallus begins to constrict. There is no clear separation of two plates in contrast to the other Melitaeinine taxa with teeth-bearing plates. The unique plate of G. sodialis is broad and triangular with the posterior side perpendicular to the plane of the phallus and the ventral side appearing almost as a continuation of the phallus. Four teeth are present, with three running approximately perpendicular to the posterior edge of the phallus and a small fourth tooth (on the right side only, based on the one specimen examined) ventroposterior to the second posterior-most tooth and near the posterior edge of the plate. I do not know if the single teeth-bearing plate attached to the vesica is homologous to the distal or proximal plates of other Melitaeini, or if the plates are so fused together that they are no longer individually detectable. If the teeth bearing plate is homologous to the distal plate of other Melitaeini, G. sodialis actually underwent a loss of the basal plate and subsequent expansion of the distal plate to reconnect with the posterior end of the phallus. This would represent an unique reversal within the Melitaeini, and an






65


autapomorphic state for the form of the distal plate. If the plate in G. sodialis is homologous to the proximal plate of other Melitaeini, then it would share a reversal with G. mundina and have an autapomorphic state for the form of the distal plate. If the apparent single plate in G. sodialis is actually the extreme fusion of both plates, then G. sodialis has an autapomorphic state for the distal plate, and does not share a reversal with G. mundina. Since these three possible scenarios have considerably different implications for coding, and I have no objective basis for favoring one of them, I code G. sodialis as "?" for the next three characters. 72. For taxa with state 2 of the preceding character, the relative size of the teeth on the distal plate.

l=Minute, all of a similar size range (Figs. 52 & 55-56).

2=Prominent, some teeth distinctly larger than others (relative to state 1) (Figs.

70,78,83,84,97,105,125 & 211-212).

While the size difference between state I and 2 is quantitative, the difference is sufficiently obvious that quantification is not necessary to assign two distinct states (see the above referenced figures). However, note that within state 2 taxa, larger species tend to have larger teeth than smaller ones, but none were found to be ambiguous in coding as distinct from state 1.

Didymaeformia didyma lacks a second plate on the vesica but has minute teeth

(identical to other state I taxa) on the vesica in the area where the second plate would be. This taxon is coded as state 1. Independent evidence from other characters suggests the second teeth bearing plate may have underwent reversal. All taxa with the distal teeth bearing plate have a proximal plate to which it is attached (although this could not be






66


interpreted in one taxon, G. sodialis). Taxa with a proximal plate but lacking the second teeth bearing plate are coded "?", while taxa lacking either plate are coded "0". 73. For those taxa with state 2 of the preceding character, the arrangement of teeth on the teeth-bearing plate attached to the vesica:

l=Teeth in one row along an arch (Figs. 105,125 & 211-212).

2=Teeth are not confined to one row (Figs. 52,55-56,70,78,83-84 & 97).

74. For those taxa with a plate bearing teeth attached to the vesica, characteristics of this plate (these descriptions are based on a view with the vesica everted):

I=A narrow, heavily sclerotized crescent curving posteriorly and bearing teeth

(usually 6-10) projecting posteriorly along the crescent (ridges of the anterior base

of the teeth occur on the outer side of the crescent) (Fig. 211).

2=As in state I except dorsally the plate extends as an anteriorly curving arc (this state occurs in only 3 taxa of the genus Chlosyne; all other Chlosyne have state 1)

(Fig. 212).

3=Base rectangular, on posterior end a ventrally projecting, anteriorly curved, and

heavily sclerotized crescent bearing teeth as in state 1 (Fig. 105).

4=Elongate and more lightly sclerotized relative to the basal plate with an

outward arch at the very base, covered with many minute teeth on surface and

edges (Figs. 52 & 55-56).

5=Narrow crescent with prominent teeth on its surface, six teeth long anterior to

posterior and no more than three teeth wide (Fig. 70).






67


6=Arched plate with prominent sclerotized teeth spread over outer surface (much

less elongate and more heavily sclerotized than state 4, with many fewer but

larger teeth) (Figs. 78,83-84 & 97).

Taxa lacking a pair of plates attached to the vesica and posterior phallus are coded "0". Taxa with such plates but lacking a second teeth bearing plate are coded "?". 75. A laterally flattened, sclerotized plate attached to the very posterior edge of the dorsal midline of the phallus, and attached to the vesica for the remainder of its ventral edge:

I =Absent.

2=Present (Fig. 78).

This appears to be what Higgins (1981) referred to as an "ostium keel." 76. For those taxa with state 2 of the preceding character, basic shape of this plate (viewed when vesica everted):

I =Triangular dorsal projection on proximal side (not illustrated).

2=Proximal side entire (Fig. 78).

State 2 of the preceding character occurs only in the Phycioditi, but there is no evidence to conclude it is a terminal derived state. Consequently, Phycioditi lacking state 2 of the preceding character are coded "?". However, as independent characters provide clear evidence the Phycioditi are a monophyletic group, other Melitaeini taxa can be coded

14011.

77. Posterior dorsal side of phallus "hinged" and changes its orientation from posterior to dorsal posterior when the vesica is everted:

0=This feature absent.






68


l=This feature present (Fig. 105).

78. The dorsal side of the everted vesica with a prominent, three dimensional sclerotized structure attached to the dorsal posterior edge of the phallus. The structure is entire with an anteriorly directed lobe proximally and elongate and tapering distally, extending well past the teeth-bearing plates of the everted vesica:

0=Absent.

1 =Present (Figs. 47-48 & 55).


Female genitalia

Characters of the Corpus Bursae:

79. The presence of a sclerotized plate on the ventral posterior side of the corpus bursae.

l=Present (Figs. 261-265,267-272,274,276-284 & 286-292).

#2=Absent.

Note that in Mellicta asteria the plate is very lightly sclerotized relative to other state I taxa, and may require higher magnification and more light to see. Since I only examined one specimen of this taxon, this feature may not be consistent. 80. For those taxa with character 79 state 1, extensions of plate on ventral posterior side of corpus bursae:

I =Plate extends anteriorly in two, narrow extensions, one on each side of the

plate (Figs. 261-265,278-284 & 286-292).

2=Plate lacks a pair of anterior extensions (Figs. 267,269-272,274 & 276).

The extensions are atypically short in Mellicta alatauica; however, the posterior arch of the plate is very characteristic of other Mellicta species examined.






69


Independent evidence suggests the presence of a plate on the ventral posterior

corpus bursae is a terminal derived state for all Melitaeini, and the out group is coded "0" for this and the next set of characters. 81. For those taxa with character 79 state 1, the posterior curve of the corpus bursae plate at the plates anterior end:

l=Anterior edge not curved posteriorly.

2=At least part of anterior edge distinctly curved posteriorly (Figs. 269-272).

82. For those taxa with character 79 state 1, the separation of plate on ventral posterior side of corpus bursae:

1 =Plate distinctly separated by membranous tissue along ventral midline (Fig.

261).

2=Plate with continuous sclerotization along ventral midline (Figs. 262-265,267272,274,276,278,279-284 & 286-292).

83. For those taxa with character 79 state 1, a raised sclerotized ridge (highest in middle) along midline of dorsal (inner) surface of the sclerotized plate on the ventral posterior side of the corpus bursae.

I =Absent.

2=Present (Fig. 269).

84. For those taxa with character 79 state 1, the presence of many small sclerotized teeth on the inner surface of the plate on the ventral corpus bursae:

1 =Present.

2=Absent.






70


The size of these teeth are too small to effectively illustrate with camera lucida drawings. although in some of the larger specimens I have illustrated a jagged edge along parts of the plate folded such that the teeth are visible along the edge (Fig. 261, for example). Taxa lacking character 42 state 1 are coded 0. 85. For those taxa with character 80 state 2, the shape of the plate on the ventral corpus bursae:

l=Very long and narrow, well over 5 times as long as the width of the anterior

side, posterior side widest and slightly wider than anterior side (Fig. 267).

2=Rectangular and wider than long, posterior side about as wide as anterior side

(Fig. 276).

3=Stout and bulged, at most 3 times as long as wide (usually less) (Figs. 269272).

4=Wider than long, anterior side about 1 .5-2X the length of the posterior side,

posterior edge irregular, sides widen posterior to anterior in a concave curve,

anterior edge extends along the midline (Fig. 274).

Independent evidence does not suggest state 2 of the fourth preceding character is a terminal derived state, so taxa lacking this state are coded "?" Out group taxa (with no plate on the ventral corpus bursae) are coded "0".

There is considerable variation within state 3, a state particular to the Phycioditi, which may potentially provide useful information for a detailed phylogenetic study of the Phycioditi.

86. For those taxa with character 80 state 1, the extent of the sclerotized plate on the ventral posterior side of corpus bursae before it diverges posteriorly:






71


1 =The unforked posterior portion of the plate extends for notably less than /2 the

length of the plate (Figs. 261,265,278-284 & 286-290)).

2=The unforked posterior portion of the plate extends for nearly 2 the length of

the plate or more (Figs. 262-264).

Note that Didymaeformia trivia has the plate a single nondivergent piece for nearly its entire length, before forking only slightly at the anterior most end.

Outgroup taxa are coded "0" and in group taxa lacking character 80 state I are coded "?".

The next four characters refer to aspects of variation in the signa. Melitaeinine signa are composed of aggregations of small teeth on the inner side of the corpus bursae. Some out group taxa have signa while others do not, and none of those that do have them in the same patterns found in the Melitaeini. The presence of signa teeth is obviously not a terminal derived state, so taxa lacking signa must be coded as "?" for characters pertaining to signa variation.

87. Aggregations of small teeth on inner surface of corpus bursae:

I =Absent.

2=Present.

88. For those taxa with character 88 state 2, the magnification required for the teeth of the signa to be readily apparent with a fiber optic light source:

l=Readily apparent in a spot plate at 25X1.5 magnification.

2=Extremely minute and barely if at all detectable in a spot plate at 25X1.5

magnification.






72

The signa teeth are easier to detect, and detectable at a lower magnification, when the genitalia are placed in a glass petri dish with water and a very bright light source from underneath. There were some cases where scoring this character within the Phycioditi was somewhat ambiguous, and examination of additional Phycioditi may warrant the conclusion that this character should not be used in a phylogenetic study of that group. 89. For those taxa with character 88 state 2, the arrangement of the aggregations of small teeth on the inner surface of the corpus bursae:

l=Continous band (of similar width) of teeth encircles corpus bursae (Fig. 261).

2=Distinct signa patches of teeth present on lateral sides, but connected by a narrower bridge of teeth on the ventral side (in Melitaea cinxia and Mellicta

britomartis the bridge of teeth is broken at the ventral midline) (Figs. 262-265).

3=Teeth confined to two distinct patches on the lateral sides of the corpus bursae

(Figs. 269-270,278,280,282,284,286,287,290 & 291).

4=A continuous band of teeth which is widest on the lateral sides, narrow across

the dorsal side, and relatively wider (than dorsally) across the ventral side (Fig.

267).

5=Nearly continuous band of irregular width encircles the corpus bursae, except for a narrow strip along the ventral midline which is nearly devoid of teeth (Figs.

274-276).

#6=The teeth densely cover the entire corpus bursae anterior of the ductus bursae. 90. For those taxa with character 88 state 2, the relative sclerotization of the teeth on the corpus bursae:






73


0=Appearing non-sclerotized to very lightly sclerotized (Figs. 261-265,267270,274,276,278,280 & 282).

1 =Prominently sclerotized (Figs. 284,286 & 289-291).

The amount of time the preparation is left in KOH has little effect on the ability to score this character.

91. Corpus bursae with a distinct protruding bulge on each anterior lateral side (signa patches of teeth cover the surface of the bulge in taxa which possess this character):

I =Absent.

2=Present (Fig. 270).

92. Ventral corpus bursae anterior to the ventral plate relative to the dorsal side.

O=Bulged out no more than 3X the distance on the ventral side relative to the

dorsal side (Figs. 268 & 291-292).

1 =Bulged out over I OX the distance on the ventral side relative to the dorsal side

(Fig. 265).

Note that this character must be scored in lateral view. Ductus Bursae:

93. A distinct sclerotized plate (originating at the posterior end of ductus bursae at the junction with the ventral genital opening) on the ventral ductus bursae separate from and posterior to the plate on the ventral corpus bursae (the plate on the corpus bursae would not have to be present to score this character due to the difference in position of the two plates).

l=Present (Fig. 261).

2=Absent.






74


Ostium Bursae:

94. A distinct sclerotized tube (ostium bursae) extending distal of the plates of lamella postvaginallis and lamella antevaginallis.

1 =Absent.

2=Present (Figs. 269-270 & 272).

All Phycioditi have state I and no other Melitaeini do. There is considerable variation in the shape of the ostium bursae within the Phycioditi, which may provide useful phylogenetic information for a phylogenetic study of that group. Lamellae and associated structures:

95. A membranous (stains blue in chlorozol black) pouch with many longitudinal wrinkle lines on its ventral surface, connected from the posterior lateral sides of the lamella postvaginallis around the anterior end of the lamella antevaginallis, and extending posteriorly over the genital opening and ostium bursae.

1 =Absent.

2=Present (Figs. 269-270 & 272).

96. Ventral lateral sides of lamella antevaginallis distinctly more sclerotized than a membranous-lightly sclerotized area in-between, giving the appearance extensions.

1 =This feature absent, sclerotization of the edge of the of lamella antevaginallis

fairly uniform.

2=This feature present (Figs. 284-291).

97. A raised posteriorly orientated hollow and compressed ridge on the lamella antevaginallis forming roughly a half circle around the ventral side of the corpus bursae opening:






75


0=Absent.

1=Present (Figs. 279-282).

98. For those taxa with state 1 of the preceding character, the form of the ridge on the ventral side of the corpus bursae opening:

1=Of similar width throughout, with the sides at most slightly produced over the

ventral midpoint (Figs. 280 & 282).

2=The sides are flared out as very prominent posterior-lateral extensions, and the

ventral midpoint of the ridge is barely produced over the lamella antevaginallis

(Figs. 279 & 281).

Independent evidence indicates state 1 of the preceding character is a terminal derived state, therefor taxa lacking this state are coded "0". 99. Development of the lamella antevaginallis relative to the lamella postvaginallis:

l=Well developed, as large as the lamella postvaginallis or only slightly smaller

(Figs. 261-263,265,274-275,277-292).

2=Very reduced, much smaller than lamella postvaginallis (Figs. 267-272). Higginsius miriam has the lamella antevaginallis much smaller than the lamella postvaginallis, yet the lamella antevaginallis is not very reduced but rather the lamella postvaginallis is unusually expanded. I code this taxon as state 1. In some out group taxa there is no distinction between lamella antevaginallis and lamella postvaginallis, rather there seems to be a single plate around the genital opening. These taxa are coded "?". 100. The formation of a pouch by the lamella postvaginallis and lamella antevaginallis around the opening to the ductus bursae:

0=Not forming a pouch, projecting in opposite or widely divergent directions.






76


1 =Forming a partial open pouch, with the separation wide enough that the

opening to the ductus bursae can be seen in ventral posterior view without prying

the plates apart with a forceps, projecting at an acute angle (Figs.

283,286,285,288,290,291 & 292-this pouch is not apparent for all taxa which

possess it at the angles from which they are illustrated).

2=Forming a closed pouch, with the opening to the ductus bursae difficult to see unless the plates are pried apart with a forceps, projecting at a small acute angle

(Figs. 262-265).

While taxa with state 0 could be easily scored, there were some cases where distinguishing between state 1 and state 2 became somewhat subjective, mainly for Antillea. There can be some variation in how much the pouch is closed within the same taxon, as was noted for some Chlosyne and Mellicta athalia. Texola tend to have the partial wider than Chlosyne. However, most individuals could be readily assigned to one of the above states. Note that Phycioditi lack any pouch formed from the lamella, however they have an analogous membranous pouch in a similar position.

Upon casual inspection, Atlantea may appear to have a closed pouch of state 2 similar to Mellicta. However, the pouch in Atlantea is formed by a posteriorly curved broad emarginate process extending off of the anterior edge of the lamella antevaginallis and fused with the lamella postvaginallis at its base. The lamella antevaginallis itself projects opposite the lamella postvaginallis typical of state 0. In state 2 taxa where the lamella antevaginallis plate is angled acutely with respect to the lamella postvaginallis, the abdominal stemite is fused to the distal end of the pouch. In Atlantea, where the






77

pouch is formed by an extra process, the abdominal sternite is attached to the base of the lamella antevaginallis plate, and the process is free of the stemite throughout its length.

The out group taxon Catacroptera cloanthe has a pouch formed by a extension of the abdominal tergite over the lamella antevaginallis. The posterior edge of the pouch formed by the sternite is at the anterior edge of the ductus bursae opening. The lamella antevaginallis itself is characteristic of state 0. 101. Given state 2 of the preceding character, the edge of the lamella postvaginallis with respect to the lamella antevaginallis at the distal edge of the pouch:

I =Extended nearly even with the lamella antevaginallis (Fig. 264).

2=Extending well posterior of the lamella antevaginallis (Figs. 262-263 & 265266).

While independent evidence suggests state 2 of the preceding character is a terminal derived state, taxa lacking state 2 of the preceding character are coded "?" because this character is not necessarily dependent on having state 2 for the preceding character. However, taxa with the lamella pouch can be readily assigned to discrete states for this character, which is not the case for taxa with the lamellae divergent. 102. An extension of the outer edges of the lamellae antevaginallis beyond the plane of the rest of the lamella antevaginallis on the side opposite the ductus bursae opening, forming concave depressions on each side:

0=This area is fairly smooth, lacking such extensions.

1 =The above extensions are present (Fig. 283).

103. Folding of the distal edge (farthest from the genital opening) of the lamella antevaginallis:






78


0=Unfolded.

I =Folded over on the inner side such that the distal edge of the lamella

antevaginallis is a three-dimensional structure (Fig. 264).

104. Sides of lamella antevaginallis near the base with a sclerotized structure with a convex outer side folded around the sides of the lamella antevaginallis and attached to the abdominal sterna ventrally and the lamella postvaginallis dorsally:

0=This structure absent.

I =This structure present (Figs. 262-266).

This and the preceding character were examined by prying the lamellae apart with a forceps as well as examining them from ventral aspect. 105. The presence of a lightly sclerotized area on the lamella postvaginallis posterior to the ductus bursae opening:

1=Absent (Fig. 261).

2=Present (Figs. 262-264,270,273-274,277-291 (several illustrations depict taxa

with this state where it can not be seen due to the angle from which the illustration

was produced).

Note that taxa found to have state 1 are also taxa with a sclerotized ductus bursae. These taxa have the distal posterior side of the ductus bursae lightly sclerotized. It may be possible that the sclerotized end of the ductus bursae in these taxa was formed from an invagination of the lamella antevaginallis and lamella postvaginallis sclerites, and that the presence of a sclerotized ductus bursae and the above character may not truly be independent.






79

106. Given state 2 of the preceding character, the extent of the lightly sclerotized area at the base of the lamella postvaginallis:

1=The entire base of the lamella postvaginallis is lightly sclerotized (Figs. 262263).

2=The more heavily sclerotized area extends to the base of the lamella

postvaginallis on the sides (Figs. 264,270,274,277-287). Taxa lacking state 2 of the preceding character are coded "T'?". 107. Given state 2 of the second preceding character, the continuity of the lightly sclerotized area at the lamella postvaginallis base with the posterior edge of the lamella postvaginallis:

1 =Narrowly contiguous (Fig. 283).

2=Not contiguous (Figs. 262-275 & 277-292).

3=Very broadly contiguous (Fig. 282).

108. Vicinity of the horizontal midline of the lamella postvaginallis:

0=Smooth, lacking a ridge (Figs. 261,266,267,269-270,273,275,277(there is a

concave depression not a ridge),278-288 & 290-291).

1=Extends ventro anteriorly as a hollow ridge (Figs. 262-264).

109. Anterior edge of lamella antevaginallis with a strongly posteriorly curved broad emarginate process overlapping the lamella postvaginallis for most of its length and enclosing the ductus bursae opening in a pouch. Dorsally the process has posterior extensions fused with the lamella postvaginallis on each side of the base which form "walls" around the genital opening.

0=Lacking a process.






80


l=With the process described above (Figs. 274-275). Ovipositor Lobes:

110. A distinct sclerotized pouch anterior to the ovipositor lobes formed from an extension of the anterior base of the ovipositor lobes on the ventral side.

#0=Absent.

1 =Present (Figs. 268,288-290 & 291-292).


Out Group Characters

The following characters provide no information pertinent to the relationships within the Melitaeini, as all Melitaeini have the same states for these characters. These characters were coded for an analysis using a smaller number of Kallimini and Nymphalini representatives as the out group (Junonia coenia, Catacroptera cloanthe, Anartia chrysopelea, A. jatrophe, A. fatima, and A. amathea). These characters may be helpful for future work on the relationships within or between these genera, but are by no means an exhaustive list. They are irrelevant to and not used in the analysis using collective out group coding.

111. Each anterior side of the vinculum projecting anteriorly in a prominent, broad and smooth-sided, heavily sclerotized anterior-ventral projection:

0=Absent.

1 =Present.

112. When viewed at a ventral posterior angle, a flattened U-shaped plate between the valves and anterior of the uncus, densely covered with numerous long, thick hairs projecting dorso posteriorly over the plate:

O=This feature absent.






81


1 =This feature present.

113. Outer posterior side of valve curved over 180' forming a short extension over the inner posterior lateral side bearing 5-8 (n=2) teeth on the ventral side and three teeth along the vicinity of midline on the surface:

O=This feature absent.

l=This feature present.

114. Outer dorsal posterior side of valvae curved over the ventral posterior inner side in a curved, flat extension tapering to a triangular point projecting inward with an anterior slant. The projection bears five (n=2) prominent teeth on its posterior edge:

O=This feature absent.

I =This feature present.

115. Posterior ventral comer of valvae extended inward in a triangular extension:

O=Absent.

1 =Present.

116. A sclerotized lateral ridge of pointed teeth on each posterior lateral side of the everted vesica, separate from the phallus:

1 =Present.

2=Absent.

117. Ventral anterior sides of phallus with a flat, flared, lateral extension slanting dorso anteriorly:

I =Present.

2=Absent.






82


118. Anterior end of phallus with a scoop-shaped membranous pouch, with the convex border of the scoop on the dorsal side and the ventral side slightly concave.

0=Absent.

1 = Present.

119. Ventral anterior side of phallus with a laterally expanded plate attached to the ventral and lateral sides of the phallus (a suture line is clearly visible). The plate is arched ventrally and U-shaped when viewed in anterior aspect:

0=Absent.

1 =Present.

120. Form of the ductus bursae:

O=Very short or vestigial.

1 =Long and tubular.

121. Form of the posterior end of the corpus bursae::

0=Corpus bursae tapers posteriorly.

I1=Tube-like and of similar width, distinctly narrower than the remainder of the

corpus bursae


Phylogenetic Analysis

A heuristic search with equally weighted characters with the representative out group method yielded 5 10 equally parsimonious trees with a consistency index of 0.85 8, a rescaled consistency index of 0.827, and a retention index of 0.965. The strict consensus tree of this analysis is presented in Figure 293. The consensus tree (Figure 293) is almost completely resolved except for relationships within the Phycioditi dlade. In fact, almost all the equally parsimonious trees are a result of permutations within the






83


Phycioditi clade. If all but two Phycioditi are deleted (or the P. tharos three species clade plus any other Phycioditi taxon), only two most parsimonious trees are obtained (before successive weighting CI=0.892, RI=0.970, RC=0.865; after successive weighting CI=0.940, RI=0.983, RC=0.924) with no change in any of the other clades. Of 115 parsimony informative characters, 88 characters require only the minimum number of steps on the most parsimonious trees, and 27 characters are homoplastic. Status of individual characters is presented in Table 2.

The strict consensus tree supports the monophyly of Higgins' (1981) Euphydryiti and Phycioditi, but indicates Melitaeiti is paraphyletic. Higgins (1981) concepts of Antillea, Higginsius, Atlantea, Eurodryas, and Hypodryas come out as monophyletic groups. However Higgins (1981) concepts of the following genera come out as nonmonophyletic (*) or make a different genus paraphyletic(#): Texola*, Dymasia#, Microtia#, Gnathotriche*, Gnathotrusia#, Occidryas*, and Euphydryas#. As noted, insufficient taxa were examined to test Higgins (1981) concepts of Phycioditi genera or Melitaea, Mellicta, Cinclidia, and Didymaeformia, however the taxa in the latter four genera have grouped together with the limited sample examined. The Chlosyne group including Chlosyne, Thessalia, Charidryas, and Anemaca, comes out as a monophyletic group.

The shortest possible trees obtainable from successive weighting with the

representative out group method were obtained from the second successive weighting analysis. Twelve equally parsimonious trees were obtained, with a consistency index of

0.926, retention index of 0.982, and rescaled consistency index of 0.910. The topology of the strict consensus tree of successively weighted characters (Figure 294) is identical to






84


that obtained from equally weighted characters, except there is greater resolution within the Phycioditi clade.

Boot strap 50% consensus trees from equally (Figure 295) and successively

weighted characters (Figure 296) with the representative out group method support the same clades with one notable exception. The clade with Chlosyne comes out as sister clade to the clade with Poladryas with a score of 51% for successively weighted characters, but this clade only has a score of 40.5% for equally weighted characters. Most clades from the two analyses have scores within 5% of each other, and only one other clade has a score with a difference greater than 9%; the clade uniting the two representatives of Cinclidia and Melitaea has a score 13% lower for equally weighted characters. None of the clades supported by the boot strap analyses are in conflict with the strict consensus trees for equally or successively weighted characters, and all of the clades appearing on the strict consensus tree of equally weighted characters appear on both 50% boot strap consensus trees with three exceptions. One exception is the missing union of the Chlosyne group and Poladryas group on the boot strap tree from equally weighted characters as noted. In addition, the relationships between the three clades coming out as the sister clade to Euphydryiti are unresolved on both boot strap trees. Finally, Mellicta varia, M. asteria, and M. aurelia came out as a resolved clade on the strict consensus trees but this clade was collapsed on both of the boot strap consensus trees.

With the cumulative out group method, the heuristic search of equally weighted characters yielded 2,546 equally parsimonious trees with a consistency index of 0.863, retention index of 0.963, and rescaled consistency index of 0.831. The strict consensus






85


tree (Figure 297) is identical to the one obtained for the representative out group method, except the three species clade of M. varia, M. aurelia, and M. asteria is collapsed. Once again, most of the most parsimonious trees are due to permutations within the Phycioditi clade, and if Phycioditi are removed except for the P. tharos three species clade and one other representative, only 10 most parsimonious trees are obtained (CI=0.9, RI=0.968, RC=0.871 with equal character weighting; CI=0.946, RI=0.983, RC=0.929 with successive character weighting). Of 108 parsimony informative characters, 83 only require the minimum number of steps on the most parsimonious trees while 25 characters are homoplastic. The status of individual characters is presented in Table 3.

The shortest possible trees derived from successive weighting were obtained after two repetitions of successive weighting. Forty most parsimonious trees were obtained, with a consistency index of 0.933, retention index of 0.983, and rescaled consistency index of 0.917. The strict consensus tree (Figure 298) is identical to the one derived from the representative out group method except for the same three collapsed Mellicta taxa noted above.

Boot strap 50% consensus trees for equally and successively weighted characters obtained with the cumulative out group method differ little with respect to their complementary trees derived from the representative out group method. The resolution obtained is identical, and the scores differ by less than 10% in every case, and the scores for nearly all clades differ by no more than 5%.

It is not possible to report exact number of character state changes on each branch length because a number of characters can be drawn on a tree in more than one equally parsimonious way. Figure 301 presents the minimum and maximum number of possible






86


state changes for each branch length, as calculated with MacClade 3.07. Figure 302 gives character numbers for those characters which undergo an unambiguous state change on each branch. The tree used to obtain this information includes the representative out group (the number of unambiguous state changes is less with the cumulative out group, because the cumulative out group is treated as one taxon) with the Junonia and Anartia clades drawn as paraphyletic. If these clades are drawn as monophyletic, there are no unambiguous state changes at the Melitaeini ancestral node, but the other values are unaffected. Switching the order of these clades in the paraphyletic out group arrangement has no affect, however. Also, since MacClade 3.07 does not calculate unambiguous changes below polytomies, even if all taxa above the polytomy node have the same state, I used the arrow under "Tools" to remove polytomies for Figs. 301-302 in an arrangement compatible with a most parsimonious tree.


Discussion

The different types of analyses including heuristic searches with equally weighted characters, heuristic searches with successively weighted characters, boot strap analyses with equally and successively weighted characters, and their application to the representative and cumulative out group methods, all support congruent tree topologies, The results of these analyses provide a well resolved and well supported phylogenetic hypothesis for a number of Melitaeinine clades, and provide a framework for detailed studies of relationships within the three clades not investigated in detail in this analysis: the genus Chlosyne, the Phycioditi, and the Eurasian taxa Higgins (1981) placed in Melitaeini. A detailed phylogenetic study of Chlosyne is presented in the following chapter.






87


One significance of the results is that they allow for the construction of a well supported phylogenetic classification of Melitaeinine subtribes based on the criteria of monophyly and stability. I find that Robbins and Henson's (1986) guidelines for applying these criteria to constructing a classification work well for Melitaeinine subtribes: "[1] If a genus [or other higher taxon] is monophyletic, do not change the name. [2] If a genus is not monophyletic, choose the combination of monophyletic generic groupings that will create the fewest name changes. [3] If another option causes more name changes now but will be more stable in the future because of better evidence of monophyly, then present the reasons and evidence for that choice" (content in brackets [] inserted by the author]. However, while I apply these three guidelines to my subtribal classification scheme, I do not adhere to a strict interpretation of the first guideline at the generic level because there are cases where synonymizing a monophyletic genus to make a paraphyletic genus monophyletic achieves the goal of monophyly with fewer name changes. Furthermore, I always choose to synonymize monotypic genera that make another genus paraphyletic. Monotypic genera are without value to reflecting evolutionary relationships, plus retaining such genera often would involve more name changes than synonymizing them and never require fewer. Consequently, I amend Robbins and Henson's (1986) guidelines 1 and 2 as follows: "If a genus is not monophyletic, or if a genus is monophyletic but makes another genus paraphyletic, choose the combination of new monophyletic generic groupings and/or generic synonymizations that will create the fewest name changes. If a genus is monotypic and comes out within a different genus, synonymize the monotypic genus (or the genus representing the clade to which it belongs, if the monotypic genus has the older name). If






88

two or more alternative natural classifications require equivalent numbers of name changes, choose the option which requires the fewest changes to valid taxonomic categories." I accept Robbins and Henson's (1986) guideline number 3 verbatim.

Applying the above guidelines leads to the retention of Euphydryiti Higgins and Phycioditi Higgins, which represent long standing taxa that come out as well supported monophyletic groups in all of the analyses. Application of the second guideline involves splitting Melitaeiti into three subtribes, with Melitaeiti now restricted to the Eurasian taxa Higgins (1981) included in his concept of Melitaeiti, Gnathotrichiti as a new subtribe forming the sister clade to Phycioditi, and a third subtribe including the Poladryas and Chlosvne clades. However, application of Robbins and Henson's (1986) third guideline favors naming the Chlosyne and Poladryas clades separately (as Chlosyniti and Poladryiti), since the evidence of the individual monophyly of these two clades is much stronger than the evidence of their combined monophyly. While placing Gnathotrichiti within the Phycioditi would require the same number of name changes and be as phylogenetically valid as giving it Gnathotrichiti subtribal status, I choose the former action because it does not require changing the boundaries of a valid taxon (Phycioditi).

Several factors can be considered when evaluating how well supported a

particular grouping of taxa is, including congruence between independent data sets, character quality, and the number of extra steps required by less parsimonious arrangements. Congruence between independent data sets is used in the following chapter, where the phylogenetic analysis includes a character set from genitalic characters plus a character set from scale pattern characters, but it cannot be used to evaluate the results presented in this chapter since all characters are based on genitalia. Character






89


quality includes simple versus complex characters, and the degree of character state uniqueness among related taxa and out groups (for example, is a character state supporting a clade a unique synapomorphy or a state that must have independently evolved more than once to account for its distribution on the most parsimonious tree(s)?). Unique character states which require only one step to account for their distribution on the most parsimonious tree(s) arguably provide greater support than individual character states which appear to have independently evolved more than once. Also, character states based on complex sclerotized structures are arguably less likely to evolve more than once (see the following chapter for empirical evidence supporting this hypothesis). I argue in the next chapter that boot strap scores may be a very poor method of evaluating support for various clades, and that clades appearing only in an analysis based on successive character weighting are of very dubious validity. In general it is logical that the greater the number of extra steps required for an alternative topology, the better supported a particular clade is. However, I hypothesize that clades supported by relatively few steps are not necessarily poorly supported if they are supported by several complex character states but with some conflicting evidence from simple homoplastic characters.

With respect to the subtribal classification which I propose for Euphydryiti, Melitaeiti, Gnathotrichiti, Phycioditi, Poladryiti, and Chlosyniti, these clades are supported by multiple complex character states (see subtribe descriptions below), and a minimum of three to twelve additional steps would be required to break apart any subtribal clade. Also, there is consistency between equally and successively weighted character analyses. A discussion of all the universal synapomorphies from binary






90

characters and terminal derived states of multistate characters is included in the subtribe descriptions below. Of course, since not all taxa in Phycioditi and Melitaeiti were examined, there is the possibility of reversals or additional character states that I am unaware of. However, given the number of characters supporting these clades, plus in the case of some characters of Phycioditi, insight provided by Higgins (1981) camera lucida drawings, I find the possibility that examination of additional taxa would significantly undermine the evidence of monophyly for these groups to be very remote.

The value of an analysis from the representative out group method in addition to the cumulative out group method is that the representative method yields fewer ambiguous state designations at the ancestral node than when the out group is collectively treated as one taxon If there is reason to suspect the taxa included in the cumulative out group and in the representative out group do not form a monophyletic sister group to the in group (this is the case for Nymphalini and Kallimini (Harvey 1991)), making the representative out group paraphyletic in MacClade is useful for investigating the number of additional steps required for alternative tree topologies.

In order to investigate the number of extra steps required for alternative

arrangements between the subtribe clades, I constructed a matrix (Table 4) including each possible pair of subtribe clades, and each possible pair of subtribe clades grouped with another subtribe clade or pair of subtribes. Values entered into the matrix represent the minimum number of extra steps required for a tree with a particular pair of clades grouped together, as determined by checking each arrangement in MacClade 3.07 from the tree obtained from equally weighted characters and with the two clades of the representative out group drawn as paraphyletic (in either order). Since the cumulative




Full Text
Table 12 Continued
732
Microtia cor acara Mexico
Morelos
to Miles S of Cuernevaca, moist forest. Station No 9 24-VIII-1967
Miller 4 Pme AM
Microtia cor acara Mexico
Morelos
Huajtndan lix.56
T Escalante AM
,
Microtia coracara Mexico
?
Mixcaie River August 1. 1942
J C Hopfinger NMNH
Microtia cor acara Mexico
?
1
VIII-46
Microtia anomalus (AM.(NMNH.FSCAJ)
' '
Microda anomakis Mexico
Michoacan
Coahuayana
VIII-50
T Escalante AM
Chlosyne harrisii (FSCA-part)
Chiosyne harris Canada
New Brunswick
Mad swaslo
Edmundston 29 June 1961
FSCA
Chlosyne harria* Canada
New Brunswick
New Brunswick
Madawaeka
Mad anraska
Edmundslon 7 July 1961
M Mensel FSCA
1
Chlosyne harnaii Canada
Edmundston 8 July 1961
M Hen sol FSCA
Chios yne harris Canada
New Brunswick
M ad v ask a
Edmundston 29 June 1961
FSCA ^
1
Chios yne harris
Canada
New Brunswick
Madam aska
Edmundslon 28 June 1963
FSCA
Chios yne harrisii
Canada
New Brunswick
Madamaska
Edmundston
Chlosyne hams*
Canada
Ontario
Port Sydney 13 June 19'0
R Michaels 1 FSCA
Chlosyne harris Canada
Quebec
Temiscouata
30 June 1963
Chlosyne harris USA
Maine
Cumberland
Bndgton ;2Jt*y197S
FSCA
i
Chlosyne harris USA
Maine
Cumberland
Portland 3 July 1972
H N Dow FSCA
Chlosyne h arris USA
Maine
Cumberland
Portland 14 June 1973
H N Dow FSCA
Chlosyne harrisii USA
Maine
Cumberland
Portland 1 7 June 1973
H N Dow iFSCA
Chlosyne harree USA
Maine
Cumberland
Portland : 14 Jtare-73
FSCA
Chlosyne harris
USA
Maine
Cumberland Portland 23 June 1975
Chlosyne harris
USA
Maine
Cumberland Portland 28 June 1975
FSCA
Chlosyne harnaa
USA
Maine
Penobscot Enfield Passadumkeag Bog
VI-26-41
FSCA
Chlosyne harrisa USA
Maine
York Shapleigh Twp N Shapleigh
25 June 1988
RGodefra
Chlosyne harris USA
Maine
? New Sueden
7/12/66
FSCA
Chlosyne harris {USA
Michigan
Jackson Waterloo Twnp Riethmiller Rd
16 June 1971
FSCA
Chlosyne harrisii USA
Michigan
Osela Lucerne
14 Jul 1976
A Godefroi FSCA
Chlosyne harris USA
New Hampshire
Coos Black Lake. Pittsburg, ex pre-pupa VI 17 83
Em VI 25 83
Chlosyne harrisa
USA
New Hampshire
Coos Black Lake, Pittsburg, ex larva 29 May 1987, Aster umbullatus
Em 9 June 1987
RW Boscoe FSCA
Chlosyne harris
USA
New Hampshire
Coos
Coos
Black Lake. Pittsburg, ex larva 29 May 1987. Aster umbullatus
Em 10 June 1987
Chlosyne harrisa
USA
New Hampshire
Black Lake Pittsburg ex larva 29 May 1987. Aster umbullatus
Chlosyne harrisii
USA
New Hampshire
Coos Black Lake. Pittsburg ex larva 29 May 1987. Aster umbullatus Em 12 June 1987
R W Boscoe
Chlosyne harris
USA
USA
New Hampshire
New Hampshire
New Hampshire
Coos
Coos
Black Lake. Pittsburg, ex larva 29 May 1989. Aster umbullatus
RW Boscoe
FSCA
2
Chlosyne harrisii
Black Lake Pittsburg ex larva 31 May 1989 Aster umbuSatus Em 10 Jue 1969
Chlosyne harrisa
USA
Coos Black Lake. Pittsburg, ex larva 31 May 1989. Aster umbullatus Em 11 June 1989
R W Boscoe
Chlosyne harrisa
USA
New Hampshire
Coos
Coos
Black Lake. Pittsburg, ex larva 31 May 1989. Aster umbullatus Em 14 June 1989 R W Boscoe
Chlosyne hams*
USA
New Hampshire
Black Lake. Pittsburg, ex larva 31 May 1989 Aster umbullatus NR R W Boscoe FSCA

Chlosyne harrisa
USA
New Hampshire
Coos Scott Bog, ex larva VIII. 12 77, Aster umbullatus Em VI 19 78 F D Fee FSCA i
Chlosyne harrisa
USA
Mew Hampshire


551
orange markings replaced by cream (except for the orange band between the position of
the submarginal and marginal bands and the hindwing postmedian band) occur in some
populations; these forms integrade to orange females. Light colored forewing markings
vary along a continuum from dark red-orange to pale cream, contrast between
postmedian and median bands variable but often with the postmedian band distinctly
darker.
Wing Span: S'- 20-41mm. $: 35-48mm.
Geographic Variation: The series in the National Museum of Natural History show a
considerable range of variation within C. acastus, but I found no indication of a gap in
the range of variation for any character and consequently recognize no subspecies.
Black and white figures of the several recently described subspecies in combination with
their descriptions also do not appear to me to fall outside the C. acastus continuum of
variation. Although the specimens I have examined included some specimens with
specific locality data and some without, it is clear that the variation of C. acastus is not a
simple cline and that different populations vary with respect to the ranges of variation in
the amount of black on the dorsal wing surfaces in females and males, the color of the
light colored dorsal wing markings, the contrast between the postmedian and median
bands on the dorsal forewing and hindwing, and the color of the ventral hindwing
background. During my visit to the NMNH collection, which has very extensive series of
C. acastus, I did not study local patterns of variation of C. acastus in detail. Based on
information from several literature sources combined with my own observations, I
attempt to summarize available information on patterns of local variation within C.
acastus below, as well as to mention how the available names apply (to phenotypes in my


685
and a few localities in western and extreme southern Brazil. I have seen no material from
most of Brazil or the Guyanas, nor any specimens from Chile. The southern most records
I have examined are Pelotas at the extreme southern tip of Brazil near the Uruguay
border, and La Roja province in northeast Argentina. Table 12 includes collection data
from the FSCA and NMNH.
Species Delimitation: Explained under C. californica (the sister taxon of C. lacinia)
above.
Note on Subspecies: Subspecies status has been assigned to forms adjutrix, crocale,
quehtala, lacinia, and saundersi, and sometimes other named forms. Since these forms
occur along a continuum of variation, they do not meet the criteria of my subspecies
concept; however, all of these forms occur also together in the same localities with at
least one other form. Consequently, to recognize any subspecies of C. lacinia, one would
need a subspecies concept where subspecies co-occur together in the same locality and
also have intermediates between them.
Chlosyne ehrenbergi (Geyer)
Figures 150,170,191,201,211,227 & 251.
Morpheis ehrenbergi Geyer, 1833. Samml. exot. Schmett. PI. 36.
Diagnosis: Chlosyne ehrenbergi is not likely to be confused with other Chlosyne due to
its highly autapomorphic wing pattern. C. ehrenbergi is the only Chlosyne with cream
colored wing fringes with no white and black checkered pattern. The dorsal forewings


217
none in fresh condition), Thessalia cynisca (but fresh representatives of both sexes were
included), Chlosyne endeis endeis (six specimens, all males), Chlosyne melitaeoides (I
examined a series of males, but only three females and two in detail), Chlosyne hylaeus
(two photographs, including a scanned image of the type (many characters consequently
had to be coded ?)), Chlosyne gciudealis wellingi (four females in detail, no males),
Chlosyne mazarum (two males), Texola anomalus (one male), and the following
Poladryiti: Higginsius fasciatus (a series of males were examined but no females),
Atlantea perezi (two males and two females), Atlantea pantoni (a series of males were
examined but only three females), Atlantea tulita (two females and three males), and
Atlantea cryptadia (I briefly viewed the holotype without associated genitalia but did not
code characters, hence this taxon was not included in the analysis).
At least three (usually more) male individuals were dissected and examined for
scoring genitalic characters, unless fewer than three males were available (for the taxa
noted above). The initial three individuals were selected from different parts of a taxons
range for those taxa not restricted to a small geographic area. In all cases where variation
was found in the first three individuals dissected, a larger series was subsequently
dissected. The vesica was everted in at least one individual of all taxa for which males
were available except for Texola anomalus, Chlosyne kendallorum and Chlosyne
mazarum. Fewer females (one to three individuals) were dissected for most taxa within
Chlosyne since no codable variation was found for the female genitalia, although there
are female genitalic differences among some taxa. More females were dissected (up to
eight) for some groups of taxa where few or no consistent differences were found


207
Strict
Chlosyniti
Poladryiti
Phycioditi
Gnathotrichiti
Melitaeiti
Euphydryiti
Figure 297: Strict consensus tree from a heuristic search of equally weighted characters
from the Melitaeini data matrix with the cumulative out group method.
Chlosyne A
Chlosyne B
T. elada
M. elva
D. dymas
T. corocara
T. anomalus
A. pelops
A. proclea
P. minuta
H. fasciatus
H. miriam
A. tulita
A. pantoni
P. tharos
P. batesi
P.phaon
P. simois
A. texana
A. tulcis
T. teletusa
C. castilla
D. catula
O. liriope
O. ithra
E. eunice
J. leucodesma
J. fellula
M. amaznica
T. claudina
T. saladellensis-
G. exclamationis
G. mundina
G. sodialis
M. athalia
M. britomartis
M. alatauica
M. parthenoides
M. varia
M. aurelia
M. asteria
C. phoebe
C. scotosia
M. cinxia
M. diamina
D. didyrna
D. trivia
E. aurinia
E. desfontaini
H. matuma
H. intermedia
H. gillettii
H. cynthia
H. iduna
O. anicia
O. chalcedonia
O. colon
E. phaeton
O. editha
Cum. Out Group


661
Chlosyne californica Wright
Synchloe californica Wright, 1905. Butt. West Coast: 170. Lectotype: California
Academy of Sciences (Tilden 1975).
Diagnosis: Chlosyne californica is quite similar to and often confused with Chlosyne
lacinia in collections. C. californica is not highly intraspecifically variable as in C.
lacinia, and resembles the adjutrix phenotype of lacinia with which it co-occurs and
more closely resembles the saundersi phenotype which occurs far south beyond the range
of C. californica. The ventral forewing discal cell is predominately orange in C.
californica and predominately black in C. lacinia (the allopatric saundersi phenotype of
lacinia has the most orange in the discal cell). The ventral hindwing and forewing
margins contain a thin row of orange scaling in the position where the submarginal band
would occur in C. californica, which is absent in all forms of C. lacinia. Ventral
hindwing cells CuA2 and Sc+Rl have a distinct orange patch in C. lacinia (except some
specimens of the saundersi phenotype), but C. californica has orange in these cells
continuous with orange in adjacent cells including the anal cells all the way to the base of
the wing (never the case in any form of lacinia). Specimens of C. lacinia within the
range of C. californica frequently but not always have the dorsal hindwing black distal to
the postmedian dots, while C. californica more often has a distinct row of orange patches
between the position of the distal parafocal elements and submarginal band. I find no
genitalic differences between C. californica and C. lacinia', however, the genitalia of


419
a: 12,25,30,39,52
57,61,62,73,75,76
77,85,93,108,129,
135,139
b: 33,34,53,55,57,69,
71,73,85,90,103,107,
108,110,116,119
*Polytomies have been removed because MacClade
does not calculate unambiguous changes below polytomies.
Refer to Figure 312 for the strict consensus tree.
38
Figure 324: Characters that change unambiguously on branch lengths on the most parsimonious tree
derived from the Chlsyniti/Poladryiti data matrix with multistate taxa interpreted as polymorphisms in PAUP.
< umPhy< inn Me


592
Austin and Smith (1998a) applied the name thekla (figures 25-30 and 37-42 in
their paper) to populations from Arizona and the name bolli (figures 31-36 and 43-48 in
their paper) to populations from Texas. I find no consistent differences between Texas
and Arizona specimens, although there is only partial overlap in the full ranges of
variation between the two states. In fact, there is a large series of specimens reared from
larvae in the FSCA from Cameron County, Texas, with identical data except for the dates
of adult emergence, which includes both specimens I can not distinguish from some of
Austin and Smith's (1998b) thekla and bolli figures plus intermediates, although the
majority of the specimens are darker and more consistent with the bolli phenotype. My
observations are that Arizona specimens on the average may tend to be lighter than Texas
specimens and have more extensive blending between the dorsal forewing median and
postmedian band, but that most of the phenotypes present in Arizona also occur in Texas
and vice versa. Likewise, I find no consistent differences in the antennae or the labial
palpi between Arizona and Texas specimens. Austin and Smith's (1998b) reference to
the amount of reddish orange on the ventral surface of the antennal shaft appears to refer
to the amount of scales present on the ventral surface, since the antennae have black and
white scales but appear somewhat reddish orange when the scales are missing. I find the
amount of scales present ventrally on the antennal shaft highly variable in all populations
of C. theona, and perhaps variable due to the condition of specimens as well (specimens
with a more worn appearance often (but not always) have fewer scales on the antennae).
Specimens from many parts of Mexico are more variable than in the United
States, and in most parts of Mexico where I have examined material, I find specimens
that exhibit the same range of variation in pattern characters, including Guerrero,


175
Figures 162-165. Male genitalia of Chlosyne sp. In dorsal aspect. 162: C. hoffmani.
163: C. acastus. 164: C. erodyle erodyle. 165: C. melitaeoides.
165
164
40X


542
B. Chlosyne whitneyi damoetas
Melitaea damoetas Skinner, 1902. Ent. News. 13:304. Type Locality: Colorado, South
Park, Hall Valley, Williams River Range. Holotype: Acad. Nat. Sci. Philadelphia
per (Higgins 1960) but in the Carnegie Mueum per (Miller and Brown 1981).
Diagnosis: Chlosyne whitneyi damoetas is the only member of the Chlosyne hoffmanni
clade which has the meso- and metathoracic femur orange and black rather than orange
and white (subspecies hoffmanni has an orange and white femur with a few proximal
black scales). The extreme diffusion of black on the dorsal wing surfaces with pale
orange markings is more subtly diagnostic, since dark specimens of C. acastus and C.
palla have a sharper contrasting black and colored pattern as opposed to a more diffuse
one.
Subspecies whitneyi is variable in the extent of diffusion of black on the dorsal
hindwing, but it never seems to be as extensive as subspecies damoetas. Subspecies
damoetas has part of the forewing more heavily diffused with black than in subspecies
whitneyi, particularly the posterior forewing basal to the distal band of the central
symmetry system and in some (but not all specimens) the area between this band and the
discal spot.
Further Description: Little contrast between different colored bands on the dorsal wing
surfaces, all are pale orange with the postmedian bands and scaling inside of the discal


546
to the wing characters (particularly the broad blending of the dorsal forewing median and
postmedian band) given for southern populations of C. hoffmanni (above).
C. gabbii should not be confused with C. whitneyi since gabbii does not occur in
alpine habitats. In contrast to C. gabbii, there is little contrast between the median and
postmedian bands in C. whitneyi and all of the dorsal markings are a pale light orange.
Also, C. gabbii has a sharp appearance to the black dorsal markings rather than the
diffused appearance characteristic of C. whitneyi, and C. whitneyi lacks any cream
scaling adjacent to the median lines.
Further Description: The pattern characters given in the further description section of C.
hoffmanni apply to C. gabbii (and C. acastus as well), except for those mentioned in the
diagnosis sections for C. gabbii and C. hoffmanni (above) and the geographic variation of
C. hoffmanni (quite unlike C. gabbii). Hence, the below description of C. gabbii requires
reference only to the color of certain wing pattern elements.
C. gabbii is far less variable than C. acastus and C. palla. The darkest orange
dorsal markings occur in the discal spot and the discal cell patches on the forewing and
hindwing where the color is a dark orange-brown distinctly darker than the postmedian
bands. The forewing and hindwing postmedian bands are orange and the postmedian
dots (cream, often with at least partial black borders in some or all of the hindwing cells)
provide a distinct contrast. The narrow orange band between the marginal and
submarginal bands is usually a slightly darker orange than the postmedian bands. This
band is always prominent on the dorsal hindwing and may be on the dorsal forewing
(most females, some males) or largely obscured by black (many males). The light
patches or crescents between the distal parafocal element and submarginal band are


403
Numbers refer to
character numbers
from Table 5
b: Basal band of basal symmetry system
c: Distal band of basal symmetry system
d: Proximal band of central symmetry system
e: Discal spot
f: Distal band of central symmetry system
g: Basal parafocal element
h: Postmedian dot (=border ocellus)
i: Distal parafocal element
Terminology not in Nijhout (1991)
j: Submarginal band
k: Marginal band
MB: Median Band
BML: Basal Median Line
DML: Distal Median Line
PMB: Postmedian Band
B: Background
(b-k are based on Nijhout 1991)
Figure 308: The Nymphalid Ground Plan wing pattern elements on the dorsal surface of a derived
Chlosyne, C. melitaeoides, and other wing pattern terminology.


510
Chlosyne nycteis (Doubleday)
Figures 134,158,176,195,213 & 234
Melitaea nycteis Doubleday, 1847. Gen. Dium. Lep. PI. 23, fig. 3. Type Locality:
Middle States. Holotype: British Museum (Higgins 1960).
=Melitaea oenone Scudder, 1862. Proc. Essex Inst. 3:166. Type Locality:
"Massachusetts, Maine" (Higgins 1960).
=Phyciodes nycteis var. drusius W. H. Edwards, 1884. Papilio 4:57. Type Locality:
California and Arizona, restricted to Turkey Creek Junction, Jefferson Co.,
Colorado by Brown (1966). Lectotype: Carnegie
Museum, Pittsburgh (Brown 1966).
=Phyciodes nycteis form lacteus Gunder, 1928. Caad. Ent. 60:166. Type Locality: St
Louis, Mo.
-Phyciodes nycteis form greyi Field, 1934. Caad. Ent. 66:256. Type Locality:
Lawrence, Kansas.
=Phyciodes nycteis form hewitsoni Field, 1936. J. Ent. Zool. 28: 23. Type
Locality: Arizona.
=Melitaea nycteis reversa F. H. and R. L. Chermok, 1940. Caad. Ent. 72:83. Type
Locality: Riding Mts., Manitoba. Holotype: Canadian National Collection (Miller
and Brown 1981).


656
and distal parafocal element, but this cream scaling is usually not detectable without
magnification. Ventral hindwing with cream-yellow patches in cells CuAl-Sc+Rl
(rarely absent in CuAl) between the position of the submarginal band and the distal
parafocal element. The size of these patches is quite variable among individuals. Red
orange postmedian band patches are present in cells CuA2-CuAl and M2-M1 with the
largest patch in CuA2, followed by either Ml or CuAl; in a few specimens the patches in
cell CuAl or M2 are reduced to a few red-orange scales. Most specimens have a single
sharp or diffuse cream patch in ventral hindwing cell 1A+2A.
Variation: C. melanarge exhibits minor variation in the size of the light colored wing
markings, which does not appear to be geographic.
Range: I have seen specimens from the Mexican states of Oaxaca (Candelaria Loxicha),
Guerrero (Acapulco) and Guatemala (Retalhuleu, Escuintla, and Zacapa), El Salvador (4
miles N of Santiago de Maria), Nicaragua (Granada, 5 miles southeast of Granada,
Managua, 12 miles south of Managua), Costa Rica (Prov. Guanacaste 8 km northwest of
Bagates), and Colombia (no locality). Table 12 includes collection data from the NMNH
and FSCA. Higgins (1960) includes some additional records for some of the above states
and countries.
Species Delimitation: Chlosyne melanarge exhibits pattern discontinuities with all other
Chlosyne, and occurs sympatrically within the range of its sister taxon, C. erodyle. Both
taxa are known from Retalhuleu and Escuintla, Guatemala.
Chlosyne eitmeda (Godman and Salvin)
Figures 144,193,208,224 & 245.


611
and some white is visible from the ventral edges of the lateral white bands; inner lateral
side with white scales and hairs dorsally and at base and orange scales and hairs ventrally
and at tip. Vertex with a distinct white centered patch reaching the posterior edge of the
plate, only black scales elsewhere or with a variable mosaic of orange scales interspersed
within the black. Sutures at the lateral edges of the vertex with black, white, and orange
scales in some individuals but lacking orange scales in others. Frontoclypeus with black,
orange, and white scales and hairs in a variable pattern; some specimens with white
scales coalesced (to varying degrees) into a white stripe along each eye dorsally, and a
white patch in the center of the ventral side of the face (sometimes fused with the white
stripes along the eyes); some specimens tend to have orange scales dominating over black
ventrally and blending to a predominance of orange scales dorsally. Black and white but
no orange scales present in the collar between the head and pronotum. Posterior edges of
posterior abdominal segments with bands of white scales on dorsal and lateral sides, only
black scales or black and orange scales anterior to these bands. Femur of meso- and
metathoracic legs with predominately orange scales dorsally and white scales ventrally
except at distal end where only orange scales are present or covered with only orange
scales, tibia and tarsi with orange scales only. Wing pattern characters not common to all
four subspecies are variable geographically and discussed below.
Geographic Variation: C. leanira leanira is by far the most variable of the four C.
leanira subspecies, and fifteen names (not counting aberrations) are available for the
continuum of variation exhibited by this taxon. Many authors have ranked C. leanira
leanira as a species taxon and recognized varying numbers of subspecies within it, with
the maximum number of subspecies recognized by Austin and Smith (1998b) who


236
was found to vary somewhat, even within taxa. The original measurements and
calculated ratios obtained are presented in Table 6, and this data is displayed graphically
in Figure 310. A clear gap was found around a value of 0.8. C. acastus and taxa with
identical genitalia exhibit ratios exceeding 0.8, and all other character 15 state 1 taxa
yield ratios below 0.8. No other gaps were found, although there do appear to be
differences in the average ratios and range of variation between some other taxa. 1 code
one character based on the distinct gap that was found.
21. For those taxa with character 15 state 1 or 3, the ratio of the length of the posterior
process over the length of the posterior edge of the valve (measurements were made in
posterior-lateral view (Figures 193-194):
l=Greater than 0.8 (Figure 210).
2=Less than 0.8 (Figures *193-194, also 195-198 & 208-209).
Taxa lacking character 15 state 1 are coded ?. Taxa lacking a posterior valve process
are coded 0.
All Melitaeini have a projection on the inner side of each valve.
22. Shape of the distal 2/3 of the inner valve process (best seen in dorsal view):
0=Distinctly tapering, distal end sharply to bluntly pointed.
l=Of nearly uniform width, distal end squared off (Figures 152-153 & 172-173).
While overall this character is best scored in dorsal view, to reliably score taxa for this
character preparations must be examined in multiple aspects. Curvature of the inner
valve process varies among taxa, and a pointed process will appear squared off in aspects
where the apex is curved away from the angle of view.
23. Curve of the posterior edge of the inner valve process in ventral view:


119
Further description: Males: Inner valve process simple, curved, and entire. Ventral
valve opening terminating posterior to the vinculum.
Females: Ventral corpus bursae plate extends anteriorly as paired extensions, and bears
inverted teeth. Inverted teeth on corpus bursae confined to two lateral patches.
Atlantea Higgins
Synapomorphies from binary characters: Openings on the ventral valvae formed by
where the valve plates twist around extending distinctly anterior of the vinculum (Fig.
80). Anterior edge of lamella antevaginallis with a strongly curved broad emarginate
process overlapping the lamella postvaginallis for most of its length and enclosing the
bursae opening in a pouch (Fig. 274).
Terminal derived states of multistate characters: The form of the inner valve process
including a short, blunt, broad projection on the anterior side at the comer where the
projection abruptly changes direction (Fig. 81). Ventral valve projection with a serrate
posterior edge and one to two small teeth on the distal anterior side (Fig. 82).
Further description: Males: Valvae lacking posterior projections. Ventral surface of
juxta fairly smooth, lacking ridges or plateaus. Tegumen projections shorter than the
length of tegumen anterior to the projections.
Females: Ventral plate on corpus bursae lacking anterior extensions, and without
inverted teeth. Inverted teeth on corpus bursae arranged in a continuous band of irregular
width, encircling the corpus bursae and widest at the sides.


500
patches in cells CuA2-Ml and composed of black scales and in some cells cream colored
scales in the center, with the cream scaling often more extensive ventrally. The
postmedian dot in cell M3 almost always has cream colored scales in the center and
sometimes lacks a complete border of black scales (no other postmedian dots lack a
complete black border); the postmedian dot in cell CuAl usually has cream colored
scales also, and the dots in cells CuA2 and M2 occasionally do, but more often they are
all black; the Ml dot is solid black. No ring of light colored scaling contrasting with the
ventral hindwing postmedian band encircling the postmedian dots. Distal and basal
median lines present on the ventral hindwing surface, distal sections straight or slightly
curved distally, basal sections straight or slightly curved basally, each section generally
lining up with those in adjacent cells. Dorsal hindwing margin between the position of
the marginal and submarginal bands black with a variable amount of diffuse orange
scaling, most intense in cell CuA2 and progressively less in subsequent cells. Ventral
hindwing and forewing in the same position with a sharp orange band. Ventral hindwing
submarginal band distinct against the background. Dorsal hindwing and forewing
patches comprising the median band orange and concolorous or with the forewing
median band somewhat lighter anteriorly. Dorsal hindwing between the submarginal
band and distal parafocal element black with diffuse orange to cream scaling or narrow
orange to cream crescents in some of the cells. Prominent patches present in the same
position ventrally where they are light cream colored and often crescent or arrow head
shaped. Patches formed between the distal and basal elements of both the basal and
central symmetry systems on the ventral hindwing coalesced together with the symmetry
system black bands giving the appearance of a continuous band. The bands formed


328
0=Black.
l=Solid brown.
#3=Diffuse light brown.
#4=Dark orange (in males).
A possible problem with the coding scheme for this and the next character is that
they may not be completely independent of the variation coded for the color of the distal
and basal elements of the basal symmetry system, since many taxa have corresponding
states for these characters. However, some taxa coded above lacked a band formed
between the bands of the basal symmetry system, plus there are some taxa with different
color forms of the band formed from the basal symmetry system bands versus the above
band. Hence, the above coding scheme seems to be the best option for coding this
variation.
In males (and females other than Higginsius), the color of the other dark lines and
patches forming the wing pattern elements contrasting with the ventral hindwing
background color on the ventral hindwing exhibit the same range of color variation as for
the preceding character, and since this variation is presumed to be dependent with respect
to that coded in the previous character, no additional characters are erected for these wing
pattern elements.
139. For those taxa where some form of the ventral hindwing band in the preceding
character is present, the color of scales between the proximal and distal elements of the
central symmetry system:
0=Orange.


Table 1-Continued
130
Characters
Taxa
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19 20
Eurodryas aurinia
2
1
1
1
2
3
2
?
1
2
2
0
1
0
0
0
0
1
0
2
Eurodryas desfontaini
2
1
1
1
2
3
2
?
1
2
3
0
1
0
0
0
0
1
0
2
*Hypodryas maturna
2
1
1
1
1
2
2
?
3
3
2
0
2
0
0
0
0
1
0
2
Hypodryas intermedia
2
1
1
1
2
2
?
3
2
2
0
2
0
0
0
0
1
0
2
Hypodryas gillettii
2
1
1
1
1
1
1
1
4
2
2
0
2
0
0
0
0
1
0
2
Hypodryas cynthia
2
1
1
1
1
1
1
1
1
2
2
0
0
0
0
0
1
0
2
Hypodryas iduna
2
1
1
1
1
1
1
1
1
2
2
0
2
0
0
0
0
1
0
2
*Occidryas anicia
2
0
1
1
1
1
1
2
2
1
1
0
1
0
0
0
0
1
0
2
Occidryas chalcedona
2
0
1
1
1
1
1
2
2
1
1
0
1
0
0
0
0
1
0
2
Occidryas colon
2
0
1
1
1
1
1
2
2
1
1
0
1
0
0
0
0
1
0
2
Euphydryas phaeton
2
1
1
1
1
1
1
2
2
1
1
0
1
0
0
0
0
1
0
2
Occidryas editha
2
0
1
1
1
1
1
1
1
1
2
0
1
0
0
0
0
1
0
2
Out Group Taxa
Colobura dirce
c
c
c
c
0
0
0
o
0
o
0
0
0
0
0
0
0
1
0
1
Kallima paralekta
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
1
Anartia jatrophe
1
0
0
0
0
0
0
0
0
0
0
0
0
c
c
0
0
1
0
1
Anartia amathea
1
0
0
0
0
0
0
0
0
0
0
0
0
c
c
0
0
1
0
1
Anartia fatima
1
0
0
0
0
0
0
0
0
0
0
0
0
c
c
0
0
1
0
1
Anartia chrysopelea
1
0
0
0
0
0
0
0
0
0
0
0
0
c
c
0
0
1
0
1
Nymphalis polychloros
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
c
0
1
Hypanartia lethe
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
1
Vanessa atlanta
1
0
Nf
Nf
0
0
0
0
0
Pcf
0
0
0
0
0
0
0
c
0
1
Cynthia cardui
1
c
LC_
0
0
0
0
0
0
0
0
0
0
0
Nr
0
0
c
0
1
Hypolimnas pandarus
1
0
0
0
0
0
0
0
0
0
0
0
0
c
c
c
c
1
0
1
Amnosia decora
1
0
0
0
0
0
16
0
0
0
0
0
0
0
0
0
0
1
0
1
Araschnia levana
c
c
0
c
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
1
Araschnia prorsa
c
c
0
c
If
0
0
0
0
0
0
0
0
0
0
0
0
1
0
1
Agais urticae
1
0
0
0
0
0
0
0
0
0
0
0
0
c
c
0
0
c
0
1
Precis octavia
1
0
0
0
A\
0
0
6
0
0
0
0
0
0
0
0
0
1
0
1
Catacroptera cloanthe
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
1
Inachis io
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
1
Junonia coenia
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
1
Polygonia c-aureum
1
0
0
0
0
0
0
0
0
0
0
0
0
c
c
c
0
c
0
1
Metamorpha elissa
1
0
0
0
0
0
0
0
0
0
0
0
0
c
c
c
0
1
0
1
Siproeta epaphus
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
1
Yoma sabina
1
0
0
0
0
0
0
0
0
0
0
0
0
c
c
C
0
1
0
1
Doleschallia bisaltide
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
1
Salamis augustina
c
C
0
8
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
1
Napeocles jucunda
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
1
Rhinopalpa polynice
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
1
Vanessula milca
c c
0
c
0
0
0
0
0
0
0
0
0
0
0
0
0 1
0
1
Antanartia delius
c c
c
c
0
0
0
0
0
0
0
0
0
0
0
0
0 c
0
1
Kaniska canace
1 0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0 1
0
1
Cumulative Out Group
1,c 0,c
0,c
0,c
0
0 0
0
0
0
0
0
0
0,c
0,c
0,c
0 1,c
0
1


273
taxon as absent for a variety of characters would be a poor strategy. Due to the inability
to hypothesize homology between M. elvci wing pattern features and the wing pattern
elements of the other taxa, 1 code the wing pattern characters for M. elvci as
Some of the wing pattern characters scored below are formed from black markings in all
or some of their states. Consequently, these features cannot be detected in individuals where the
relevant part of the wing is solid black. In many taxa, there is intrataxon variation in how much
of the wings are solid black, and there are some specimens for which these characters can be
scored. If all specimens examined have the relevant position on the wing solid black, they are
scored as ?. I have specifically pointed out some but not all cases where this situation arises; in
general, the form of any black or potentially black wing pattern element is coded as ? if all
specimens examined have the wing solid black in the area where it occurs.
82. Presence of a narrow band of orange scales sticking out along the basal part of the
basal forewing costal margin.
0=Mixed orange and black scales, no distinct row of orange.
l=Present, a distinct row of orange.
2=Absent, black scales are present.
83. Forewing fringe of scales:
0=Checkered black and white, with eight white sections divided by black sections
at the ends of veins 1A+2A-R3. The white section across from cell CuA2 may be
subdivided by black hairs at the end of the veinlet.
*l=Cream colored, with a variable amount of brown scales across from cells
CuA2 and CuAl.
*2=Composed of all black scales.
#3=A mosaic of black, orange, and white scales.


643
Guerrero (Acahuizotla, Acapulco, 2 miles W of Colotilipa, and Roahuizatla), Oaxaca
(Huajuapan de Len and San Jos Pacifico), Puebla (30 miles NW of Acatlan), and
Veracruz (Catemaeo). Table 12 includes data from the NMNH, FSCA, and AM.
Scott (1986) reports Chlosyne marina melitaeoides has been caught once in south
Texas, but his specimen 213a (lacking accompanying data) identified as Chlosyne marina
melitaeoides is in fact Chlosyne marina. However, the distribution Scott (1986) reports
for Chlosyne marina marina ("central and southern Mexico) is consistent with specimens
I have examined of C. marina. Stanford and Opler (1993) show an Arizona and Texas
record for "Chlosyne marina (incl eumeda and melitaeoides)", but these records are not
useful for working out distributions since their species concept is a paraphyletic
assemblage of two or three separate species. I have seen no specimens of C. marina, C.
melitaeoides, or C. eumeda from the United States, but suspect the latter two could occur
as strays. C. melitaeoides is known from Montemorelos and San Fernando Mexico, each
only about 100 miles from the Texas border.
Species Delimitation: The species concept of Chlosyne marina has included C.
melitaeoides, C. eumeda, and C. hylaeus; however, as noted in the preceding chapter this
is inconsistent with phylogenetic evidence. Phylogenetic evidence indicates C. marina is
not most closely related to C. hylaeus and C. eumeda (see preceding chapter). There is
no evidence by which to conclude C. marina is or is not the sister species of C.
melitaeoides, but the two taxa do appear to be allopatric. C. marina differs consistently
from C. melitaeoides by several pattern characters and by forewing shape. If C. marina
and C. melitaeoides are found to be sister taxa, in the absence of information on the
possibility or viability of hybridization, the decision as to whether C. marina and C.


653
Range: I have seen specimens from Costa Rica (San Mateo, San Jose: Villa Colon, Prov.
Alajuela 6 km W Atenas, and Prov. Guanacaste: 5 km NW Canas, Nosara, and Nr. Plata
Del Cocos), Panama (Canal Area, Ft. Kobbe, Chiriqui Prov.: Santa Cruz, Paraso, and
Veraguas Ballema), and Colombia (no localities). Collection data from the NMNH and
FSCA are presented in Table 12. Higgins (1960) reports specimens labeled Venezuela
with no further data, and Colombian localities of Buenaventura, Manaure, Sta. Maria,
Valdevia, and Bogota.
Chlosyne melanarge (Bates)
Figures 145,222 & 243.
Synchloe melanarge Bates, 1864. Ent. mon. Mag. 1:85. Type Locality: Guatemala
(interior). Holotype: British Museum.
Diagnosis: Chlosyne melanarge is the only Chlosyne with black dorsal forewing surfaces
except for a cream colored forewing band extending from the distal posterior wing
comer to the costa. In collections, C. melanarge has been confused with C. hippodrome,
however in C. hippodrome the light forewing markings are a clean white, and a dark red
median band is present on the ventral hindwing (absent in C. melanarge).
Further Description: Black and white but no orange scales and hairs on the labial palpi.
Dorsally palpi with black scales and hairs only; outer lateral side white ventrally-center
and at base while black dorsally and at tip, black hairs around the edge with some white
ones mixed in on the ventral side; ventrally all black except for lateral white bands visible


209
Bootstrap
87
71
77
95
76
81
81
78
100
81
89
69
99
90
57
63
85
81
73
55
84
83
99
65
63
85
76
93
92
84
100
86
58
78
80
84
81
88
Chlosyne A
Chlosyne B
T. elaaa
M. elva
D. dymas
T. corocara
T. anomalus
A. pelops
A. proclea
P. minuta
H. fasciatus
H. miriam
A. tulita
A. pantoni
M. athalia
M. britomartis
M. alatauica
M. parthenoides
M. varia
M. aurelia
M. asteria
C.phoebe
C. scotosia
M. cinxia
M. diamina
D. didyma
D. trivia
P. tharos
P. batesi
P. phaon
P. simois
A. texana
A. tulcis
T. teletusa
C. castilla
D. catula
O. liriope
O. ithra
E. eumce
J. leucodesma
J. fellula
M. amaznica
T. claudina
T. saladellensis
G. exclamationis
G. mundina
G. sodialis
E. aurinia
E. desfontaini
H. matuma
H. intermedia
H. gillettii
H. cynthia
H. iduna
O. anicia
O. chalcedonia
O. colon
E. phaeton
O. editha
Chlosyniti
Poladryiti
Melitaeiti
Phycioditi
Gnathotrichiti
Euphydryiti
Cum. Out Group
Figure 299: Boot strap 50% consensus tree for equally weighted characters from the
Melitaeini data matrix with the cumulative out group method.


343
parsimony analysis). The UPGMA tree derived from pattern and genitalia characters
combined had 28% (14/50) of the groups conflicting with both the UPGMA trees derived
from pattern and genitalic characters separately (versus 0% for equal weighting and
parsimony), and 66% of the groups conflicting with either the pattern character tree or the
genitalic character tree (versus 0% for equal weighting and parsimony). In ten
replications of generating two random trees for 50 taxa and comparing each pair of trees,
in all replications 100% of the clades in one tree of the pair were in conflict with the other
tree.
Variation in Boot Strap Scores for Various Clades as a Result of the Number of
Taxa Included in the Analysis
The boot strap scores calculated for each node in different analyses with different
numbers of taxa are presented in Table 10, in addition to the high, low, range, mean,
standard deviation, and median for each node included in more than one analysis. The
node numbers designating the clade used in an analysis (left column) are based on the
nodes designated in Figure 312. The nodes for which boot strap scores are reported (top
row) are designated in Figure 311. Examples of the relationship between number of taxa
and boot strap score at a node are presented graphically for 29 nodes (not including all of
the nodes in Table 10 to avoid too much clutter) in Figure 336.
There was a great deal of variation in boot strap scores for the same node with
different numbers of taxa in the analysis. As the number of taxa increased, boot strap
scores decreased or increased by varying amounts, stayed about the same, or fluctuated
up and down. The least common pattern was for boot strap scores to increase as the
number of taxa increased. The range of boot strap scores for the same node varied


709
designated Acahuizotla. Some of the integrades are like marianna except with a partial
red basal patch, some are like gloriosa except with a partial red basal patch, and some are
intermediate between the gloriosa and marianna phenotypes with a partial red basal
patch. The partial red basal patch may be the same color as typical janais, or various
shades of darker red to red-brown. Hence, the various forms represent points along a
continuum of variation occurring over a nonuniform geographic gradient, and I recognize
no subspecies let alone separate species. The forms all occur together with integrades
along an apparently narrow blend zone, where intermediate forms are common. This
suggests the intermediate forms are not hybrids between different lineages, but part of a
single geographically variable lineage. In most of the range the nominate janais form
occurs, in some Mexican localities only marianna or gloriosa phenotypes appear to be
present, and the irrubescens phenotype appears to only occur in areas of integredation.
Consequently, the names of the forms except for irrubescens can be useful for describing
geographic variation and segregating distributional data, as I have done below, as long as
it is recognized these forms (=phenotypes) are not distinct and do NOT refer to separate
taxa of any rank. I have also seen intermediate specimens between the janais and
marianna/gloriosa phenotypes from the localities of "Mexeala" in the state of Guerrero
and "Coahuayana" in the state of Michoacan, Mexico. Higgins (1960) reported having
examined a series including irrubescens, typical janais, and gloriosa from the localities
of Jalisco and Tenacatita in Mexico.
Further Description: Black and white but no orange scales and hairs on the labial palpi.
Dorsally palpi with black scales and hairs only; outer lateral side white ventrally-center
and at base while black dorsally and at tip, black hairs around the edge with some white


678
except some remnant of a light median band is present in some of the anterior cells (at
least Ml-Sc+Rl). Progressing from form saunders i to pauper a, integrades undergo
partial reduction in the amount of orange on the dorsal forewing before reduction on the
hindwing begins, and reduction of orange on the hindwing begins with a separation of the
orange basal/median patch into a distinct median band and basal patch and progressive
decrease in the size of these areas. Like form saundersi, form paupera is usually
correctly determined as C. lacinia in collections I have examined.
In the United States, forms adjutrix, crocale and their integrades are present. Five
representatives of form lacinia and one example of form saundersi are present in the
NMNH labeled "Ariz. Barnes Collection", respectively. However, since the Barnes
Collection is the source of several suspicious records, I am skeptical of the validity of the
Arizona origin of these specimens. I have seen no other specimens of form saundersi
from the United States or adjacent parts of Mexico. However, there is an additional
record of form lacinia from California (San Miguel Mountains, 14 Jun 1959, J. C.
Hopfinger). Neck (1980) reports collecting a specimen in Texas resembling paupera but
does not illustrate the specimen.
Most Texas specimens examined are form adjutrix, but I have seen examples of
crocale labeled from Cautil lo, El Paso, and San Antonio (Bexar County) and
adjutrix/crocale integrades from San Antonio and Beeville. I have examined regular
examples of adjutrix labeled from all of these same localities as well. Most specimens I
have examined from New Mexico are form crocale, but I have seen adjutrix and
adjutrix/crocale integrades from Eddy and Donna Ana Counties in the southeast and
south central parts of that state, respectively. 1 have seen only form crocale from


423


624
especially considering Barnes material has been the source of a number of suspicious
unique records. All specimens 1 have examined which are labeled fulvia in the NMNH
and FSCA collections from Mexico were misdetermined specimens of subspecies cyneas,
including in the case of the former the oranger phenotype of subspecies cyneas noted
above; however, there are reports of the occurrence of subspecies fulvia from Mexico in
the literature. Tinkham (1944) reported the occurrence of Melitaea fulvia in the Mexican
state of Nuevo Leon eight miles south of Galaena. Smith and Brock (1988) and Stanford
and Opler (1993) indicate the occurrence offulvia in Chihuahua in addition to Nuevo
Leon. Higgins (1960) considered fulvia a synonym of alma, and his distribution of alma
is suggestive of a mixing of leanira and fulvia records.
Chlosyne leanira cyneas (Godman and Salvin)
Melitaea cyneas Godman and Salvin, 1878. Proc. zool. Soc. Lond. 1878:269. Type
Locality: Oaxaca, Mexico. Holotype: British Museum (Higgins 1960).
Diagnosis: In collections I have seen subspecies cyneas most often confused with
subspecies fulvia, especially the orangest cyneas specimens, and occasionally confused
with subspecies leanira. It may be separated from both of these taxa by the characters
given in the diagnosis of subspecies leanira and subspecies fulvia. Smith and Brock
(1988) appeared to be only aware of the more common dark phenotypes of C. leanira
cyneas, and their diagnosis will not separate these phenotypes from subspecies fulvia.


27
and dissolving away proteinaceous material with a 10% KOH solution. The maximum
magnification at which I examined genitalia was 75X, and light sources included fiber
optics and/or lighting from below the microscope platform. Some preparations were
stained with chlorozol back dye to highlight membranous areas.
Production of Melitaeini Genitalia Figures to Illustrate Character States
All the genitalia illustrations (Figs. 1-292) are from camera lucida drawings. I
drew these figures at 50X magnification, except for Figure 292 (drawn at 25X
magnification). All the drawings were done in pencil, with differing degrees of shading
used to show the difference between darker and lighter areas. I scanned each drawing,
retaining its original size, with a Hewlett Packard Scan Jet ADF at 200 DPI with the
output type set at grayscale. The drawings were then scanned to Adobe Photoshop 5.0.2,
where I selected Image: Adjust: Brightness/Contrasf and increased the contrast to 40%.
The drawings were then saved as TIFF files, and imported into Microsoft PowerPoint,
again with the original size of the drawing retained. Subsequent alterations in size (if
any) as indicated on the figures, were done in PowerPoint.
Phylogenetic Analysis
Eight separate analyses were conducted with PAUP version 4.0b4a for Macintosh,
four for the cumulative out group method and four for the representative out group
method. The four types of analyses included a heuristic search with equally weighted
characters, a series of heuristic searches with successively weighted characters
(weighting was based on character rescaled consistency indices) until the tree length no
longer changed, a boot strap analysis with equally weighted characters, and a boot strap


639
the width of black basal to the postmedian band patches in cells CuA2-CuAl and M2-M1
is less than the basal to distal width of these patches, while in C. enrueda the basal to
distal width of the black is greater than or equal to the basal to distal width of the
postmedian band patches in cells CuA2-CuAl, M2, and usually also Ml. The sections of
the distal median line are often continuous in C. marina, and when they are not they still
span the width of their cell, while in C. eumeda these sections are not continuous and
some or all do not span the width of their cell. All C. eumeda I have examined are
distinctly larger than C. marina, and the forewing of C. eumeda has a more elongate
shape than C. marina. The largest male of C. marina I have examined has a wing span of
36mm while the smallest C. eumeda male I have seen is 42mm. The equivalent
measurements are 42mm versus 47mm in C. marina and C. eumeda, respectively.
C. marina can be separated from C. hylaeus by some of the same characters
which distinguish it from C. eumeda, however the most obvious characters are the dorsal
forewing median band patches in cells CuA2 and CuAl broad and aligned basally and
distally as in C. eumeda, and the dorsal forewing postmedian dots cream colored like C.
eumeda rather than a sharp white. However, C. hylaeus has a dorsal forewing
postmedian dot in cell M3 like C. marina, and the postmedian band patch in that cell is
limited to sparse scaling around that dot. Also, the distal median line may be with
complete sections like C. marina or with incomplete sections like C. eumeda.
Further Description: Black but no orange scales and hairs on the labial palpi. Dorsally
palpi with black scales and hairs only; outer lateral side white ventrally-center and at base
while black dorsally and at tip, black hairs around the edge with some white ones mixed
in on the ventral side; ventrally with base white and black hairs and scales in the center


478
eleda. M. eleda ranges south into Mexico where I have seen records for the states of
Sinaloa, Hidalgo, Michoacan, Guerrero, Chiapas, and Oaxaca. Collection data from the
FSCA are presented in Table 12. Stanford and Opler (1993) report additional records
from the Mexican states of Sonora, Chihuahua, Coahuila, Tamaulipas, Nuevo Leon, and
Durango.
Remarks: Despite the great similarity of pattern characters between Microtia eleda and
Microtia dymas, there is considerable evidence from genitalic characters that M. dymas is
more closely related to M. cor acara and M. anomalus than to M. eleda. The similarities
between M. eleda and M. dymas represent symplesiomorphies with respect to their
common ancestor, and there are no characters (including homoplastic characters) to
suggest a sister taxon relationship between M. eleda and M. dymas.
Species Delimitation: M. eleda has unique male and female genitalia, a unique wing
pattern, and no sister species.
Microtia elva Bates
Figures 110-111, 116-117, 125-127 & 282.
Microtia elva Bates, 1864. Ent. Mon. Mag. 1:83. Type Locality: Guatemala (interior).
Holotype: British Museum (Higgins 1960).
=Microtia elva form horni Rebel, 1906. Verh. Zool.-bot. Ges. Wien. 56:377. Type
Locality: Oaxaca, Mexico. Holotype: Vienna (Higgins 1960).
-Microtia elva form draudti Rober, 1914. Seitz 5: 455. Type Locality: Mexico.


628
Both the dark phenotype and the oranger phenotypes which can be easily confused with
subspecies fulvia occur in both Arizona and at least the Mexican states of Veracruz and
Morelos, and I suspect throughout the cyneas range. There are no gaps in the range of
variation between the darkest and orangest phenotypes of cyneas, and hence I recognize
no additional taxa under the name cyneas.
Distribution: Most of the specimens 1 have examined are from southeastern Arizona near
the Mexican border, including the Huachuca Mountains and "Paradise" and "Palmerlee"
(Cochise County). Smith and Brock (1988) report there are historical but no recent
records from the Chiricahua Mountains, and one recent record from the Mule Mountains
north of Bisbee (all in Cochise County, Arizona). I have also seen Mexican records far
south and east of the Arizona localities, from the states of Hidalgo, Morelos
(Popocatepetl Park) and central Veracruz. Table 12 includes data for specimens from the
NMNH and FSCA. Stanford and Opler (1993) report records for the Mexican states of
Chihuahua, Coahuila, Sinaloa, and Durango, while Scott's (1986) range map seems to
include Sonora and Sinaloa and also possibly Nuevo Leon. Stanford and Opler (1993)
show records for only two U.S. counties, Cochise (the same county as for the Arizona
records I have examined) and Hidalgo County, the southwestmost New Mexican county
bordering Cochise County.
I have examined series of specimens of both subspecies fulvia and subspecies
cyneas from Cochise County, Arizona. According to Scott (1986), cyneas occurs in the
Huachuca Mountains above 1800m, while fulvia occurs in the flats around the Huachuca
Mountains. The range of cyneas appears to begin at about the same point where the
range offulvia ends, yet as noted for subspecies fulvia above, the orangest phenotypes of


258
lacinia an autapomorphic state. Specimens examined of the forms quehtala and lacinia
were found to have a consistent pattern, but which is also unlike that found in other taxa.
Texola anmalas is coded as ? for this character.
There are some noteworthy patterns to mention found in some of the taxa with
both orange and white scales on the front of the face between the eyes, even though this
variation varied too much intraspecifically to code it into discrete states. Some but not all
specimens of Chlosyne gabbi, C. acastus, C. palla, C. hoffmani, C. leanira fulvia, and
Chlosyne leanira leanira had a white stripe along each eye dorsally and a white patch in
the center of the ventral side of the face (sometimes fused with stripes along eyes). These
features showed varying degrees of development intraspecifically. Still other specimens
of Chlosyne acastus had these features in addition to a white stripe below and centered
between the antennal bases, with intermediates between this and the previous form.
Specimens of Chlosyne theona, C. chinatiensis, and C. definita had this feature as well,
but a discrete state could not be assigned due to the continuum of transition between this
state and no distinct white patches at all, as found in C. acastus. C. endeis specimens had
white triangular patches below the eyes with predominantly black but scattered orange
scales and hairs elsewhere, and these specimens lacked a white stripe. Finally, C. leanira
leanira tended to have a predominance of orange scales on the ventral side of the face
blending to predominantly black dorsally, as well as a prominent white patch at the
ventral midpoint of the face.
66. For those taxa lacking orange scales and hairs on the frontoclypeus, the pattern on
the frontoclypeus between the eyes laterally and between the antennae and palpi base


559
with the dark hairs not extending to the tip and the base with only white scales and hairs;
inner lateral side with orange scales and hairs ventrally and at tip and white scales and
hairs dorsally and at base. Vertex with a distinct white centered patch reaching the
posterior edge of the plate, black and orange scales elsewhere. Sutures at the lateral
edges of the vertex with black, orange, and white scales. Tuft of white hairs on the
frontoclypeus anterior-lateral of each antennal base well developed. Black, orange, and
white scales present on the frontoclypeus, sometimes with the white scales coalesced into
a mid ventral patch and/or stripes below the antennal bases. Patch of orange scales at the
edge of the inner anterior-lateral margin of the eye. Black, white, and orange scales
present in the collar between the head and pronotum. Posterior edges of posterior
abdominal segments with bands of white scales on dorsal and lateral sides, black and
orange scales anterior to these bands. Ventrally abdomen with two longitudinal parallel
black stripes against a light background, often appearing broken between each segment.
Tibia and tarsi of prothoracic legs with orange scales dorsally and white scales and hairs
ventrally. Femur of meso- and metathoracic legs with orange scales dorsally and at distal
end and white scales ventrally, tibia and tarsi covered with orange scales only. Meso-
and metapleurons where the femur tucks in with black scales and hairs and sometimes
light orange ones, white scales underneath the dark ones. Thorax ventral to the wings
with predominately white scales and hairs in females except orange and black scales/hairs
concentrated around the coxae, males with a mix of black and white scales and hairs.
Antennal shaft with a checkered pattern of black and white scales with or without orange
scales, club with a white patch on the outer lateral side.


571
Mountains, Presidio County, Texas, 4500' elevation. Lectotype (designated by
Austin and Smith (1998a): California Academy of Sciences.
Diagnosis: Chlosyne chinatiensis can be confused with C. theona and C. leanira fulvia.
C. chinatiensis is also similar to C. perlula, but since 1 have seen C. perlula only in South
America and C. chinatiensis only from Texas and Mexico, the two taxa should not be
confused.
In C. leanira fulvia and all C. leanira subspecies, the dorsal hindwing row of
postmedian patches (small and appearing as dots in some specimens) are cream colored,
while this position is orange and continuous with adjacent orange in C. chinatiensis.
Also, all C. leanira subspecies lack orange inside of the ventral hindwing postmedian
band while orange patches occur within this band in C. chinatiensis.
In C. chinatiensis there is little (if the color is orange and cream) or no (if the
color is only orange) distinction between a median and postmedian band on the dorsal
hindwing surface, the area is orange and continuous with adjacent orange. In C. theona.
both the dorsal hindwing median and postmedian bands are bordered by black basally
and distally and they are not even partially concolorous, although the color of each is
variable. The ventral forewing of C. chinatiensis basal to the distal parafocal elements is
predominately washed out orange with very little black contrast, while prominent black
contrast is present in all variants of C. theona.
In C. perlula, dorsally and ventrally the forewing has more contrasts, with a
black band dividing the median band between cells CuAl and M3 and a much more
contrasted larger pale cream postmedian band compared with C. chinatiensis. Also, in C.


590
Regardless of how dark specimens are dorsally, the discal cell is always predominately
orange ventrally, and the orange often spills into adjacent cells, but otherwise dark
specimens have black in most of the same positions on dorsal and ventral forewing
surfaces. The veins are at least thinly bordered with black except around the discal cell.
The orange in the costal cell is continuous with adjacent orange and does not form a
distinct patch, even in the darkest specimens, but some specimens have pale whitish
cream scaling along the veins in the anterior end of the discal cell.
The dorsal hindwing basal area varies from predominately orange to solid black
with only one small orange patch in the discal cell (between the bands of the central
symmetry system in the distal discal cell fork). When orange dominates, sharp or diffuse
bands of the central symmetry system are detectable in the same position where they
occur dorsally. The median band (cells 1A+2A through Sc+Rl) and postmedian band
(cells 1A+2A through R5 or Sc+Rl) vary in color and width as with the respective
forewing bands, but are always distinctly separated by black except in cell 1A+2A. The
presence of distinct or diffuse cream patches distal to the postmedian band is highly
variable, but if sharp patches are present they are small and most often occur in cells
CuAl-M2 but occasionally in cells CuA2 through Sc+Rl.
The ventral hindwing is quite similar to C. chinatiensis, with heavy black scaling
along the veins (except where they cross the orange area between the bands of the central
symmetry system) against a cream background. The central symmetry system bands are
black with orange between them, and are often confined to cells M2-Sc+Rl but
sometimes extend to cell CuA2. In some specimens the orange extends basally along the
veins all the way to the hindwing base. The parafocal elements are fused into a


295
position of the postmedian band being black it is impossible to know how many cells the
band occupies.
Some Chlosyne have a postmedian band of light colored patches on the forewing
in the equivalent position as the postmedian band on the hindwing. However, in marked
contrast to the hindwing postmedian band, the presence/absence of a forewing
postmedian band is a highly intraspecifically variable character in a number of the taxa
which possess it, varying along a continuum between completely absent to well
developed. Consequently, I do not code any characters for the forewing postmedian band
of light colored patches.
All taxa which have an upperside hindwing postmedian band also have an
underside postmedian band. However, the forms of this band on the dorsal and ventral
surfaces is notably different in many taxa.
102. Cells occupied by the ventral hindwing postmedian band:
0=CuA2-Sc+Rl (Sc+Rl may contain a black patch).
l=CuA2, CuAl, M2, and Ml.
2=CuA2-Ml.
#3=CuA2-R5.
*4=None.
The hindwing postmedian band can be difficult to define on the pale washed out
undersides of Higginsius species, but state 3 appears to be appropriate based on some of
the freshest specimens (these are also the cells occupied by the hindwing postmedian
dots).


760
necessarily less accurate indicators of phylogeny. On the average, pattern characters
were shown to have notably more instances of homoplasy relative to genitalic characters,
yet there were no cases of conflict between clades derived from analysis of pattern and
genitalic characters separately, and a number of the same clades were resolved in both
data sets. This is consistent with Robbins (1991) findings in a phylogenetic study of the
seven species of Rekoa Kaye (Lepidoptera: Lycaenidae), where separate analysis of
quantitative characters yielded a lower consistency index compared to a separate analysis
with qualitative characters, but the tree topologies were identical. Wiens (1999) argued
that a data set with more characters is preferable to one with fewer, even if the additional
characters increase the proportion of homoplasy in the data set, and the findings of this
case study combined with those of Robbins (1991) provide evidence in support of this
hypothesis.
As a result of running multiple boot strap analyses with different sized
monophyletic groups and comparing the value of boot strap scores obtained for the same
clades, I have to seriously question the utility of boot strap scores as an index of support
for particular clades. The observation that boot strap scores for particular clades may
greatly decrease, greatly increase, or remain fairly constant based on the size of the
monophyletic group included in the analysis is not consistent with a hypothesis that these
values distinguish between better and more poorly supported clades. Relative values of
boot strap scores varied all over the board with different sized monophyletic groups
included in the analysis.
Finally, an interesting question which I sought to investigate with the
Chlosyniti/Poladryiti as a case study is: do the phylogenetic methods employed in this


12
The testable hypotheses on which species delimitation are based include the
following: 1) Is the species taxon monophyletic?; and 2) is the species taxon biologically
capable of reticulating with another evolutionary lineage (present or future)?. The first
hypothesis is tested by cladistic analysis. Evidence that I find that substantially supports
(although not necessarily guarantees) the second hypothesis includes any one or a
combination of the following: 1) Sympatric occurrence of two distinct taxa; 2)
Discontinuity in the range of morphological variation of a sclerotized structure, especially
in the structure of the genitalia; 3) A lineage has no sister species, and is sympatric with
at least one of the taxa in its sister clade; or 4) Evidence of hybrid inviability with closely
related lineages. In these cases, the evolutionary species concept is applied easily in
actual practice, because the evidence supporting the testable hypotheses of an
evolutionary species is fairly clear.
However, now consider a monophyletic group of two allopatric taxa with
identical genitalia, no known information regarding hybrid inviability, but with
morphological discontinuities in wing pattern or other pattern characters. Based on this
evidence, these taxa may be hypothesized as distinct evolutionary lineages, but it is
impossible to know whether or not they are evolutionary species. This is where I feel a
subspecies concept can be useful. Theoretically, I define the subspecies category as
follows: A subspecies is a single lineage of ancestral descendent populations of
organisms that currently maintains its identity from other such lineages but that has not
lost its biological ability to reticulate with another lineage except by geographic isolation.
The future fate of a subspecies could be a reticulation event with another subspecies or
further divergence into a separate species. In practical application, I delimit subspecies


485
always absent on the antennal shaft (present in some specimens of M. eleda). Ventral
hindwing between the position of the marginal and submarginal bands black with at most
diffuse orange scaling (versus a distinct orange band in M. elecla). Relative to M. eleda,
the proximal band of the central symmetry system in cells CuA2-M3 is more often
diffuse or absent and consequently there usually (but not always) does not appear to be an
additional band of orange patches basal to the median band.
Geographic Variation: I find no gaps in the geographic variation of any character, hence
I recognize no subspecies. Scott (1986) states males have more black on the upperside in
subspecies chara of California and Arizona compared to subspecies dymas of New
Mexico and Texas. Miller and Brown (1981) listed chara as a separate species from
dymas. Series which I have observed are variable in the amount of black on the
upperside even from the same locality. I have examined several specimens is where the
dorsal hindwing postmedian band is almost solid black with only sparse diffuse scaling,
but this phenotype and intermediate phenotypes between this condition and a normal
orange postmedian band do not appear to be confined to any subset of the species' range.
Higgins (1960) argued against retaining dymas and chara as distinct local forms, and I
have made no observations to contradict this view, although the average amount of black
on males may increase in western populations, and in populations from Mexico. Higgins
(1960) reported specimens he had seen from Palm Springs had a tendency on the upper
surface to increased dusky suffusion in the spring brood and paler specimens in the
summer brood, with less color contrast. Specimens from nr. Palm Springs in the
FSCA collected on March 29 1960 include pale less contrasted individuals and darker
individuals, as do specimens labeled from the same locality on Sept. 12, 1965.


352
with the entire data set indicate the best hypothesis is that C. eumeda and C. hylaeus are
sister taxa and this clade is more closely related to the clade including C. lacinia, C.
ehrenbergi, and C.janais than to C. marina or C. melitaeoides. Also, three or eight extra
steps (with PAUPs treatment of multistate taxa and the distinct state model, respectively)
would be required to place C. hylaeus and C. marina together as sister taxa, and all
analyses support the hypothesis that C. hylaeus is most closely related to C. eumeda.
Consequently, inclusion of C. eumeda in the same species taxon as C. marina and C.
melitaeoides is highly unjustified, as is the inclusion of C. marina and C. hylaeus in the
same taxon. The similarities in wing pattern shared by C. eumeda, C. melitaeoides, and
C. marina are symplesiomorphies at the level of universality being considered. Likewise,
similarities in wing pattern between C. marina and C. hylaeus (none of which are unique
to these two taxa) are symplesiomorphies at the level of universality being considered.
The results are inconclusive as to whether C. melitaeoides and C. marina are
sister taxa (their positions as sister taxa or adjacent taxa are equally parsimonious with a
regular PAUP analysis or the distinct state model). While these taxa are allopatric, since
there is no evidence to suggest they are sister taxa, including them as members of the
same species taxon is unjustified.
The placement of C, eumeda and C. hylaeus as sister taxa is poorly supported, and
two alternative topologies require only one extra step: the placement of C. hylaeus as the
basal taxa in the (melanarge{erodyle,poecile)) clade and the placement of C. hylaeus as
basal to C. eumeda. Careful examination of a series of specimens of C. hylaeus is needed
to better understand the variability of this taxon, its character states, and consequently the
relationships of C. hylaeus with C. eumeda and other Chlosyne.


451
18.Ventral hindwing background color cream Chlosynepalla
18. Ventral hindwing background color white to pearly white 19
19. Dorsal surfaces pale orange with heavy black suffusion and little contrast between
any of the light dorsal patches Chlosyne whitneyi whitneyi
19. Dorsal wing surfaces lacking the above combination of characters
Chlosyne acastus
20. Ventral hindwing with a thick area of black scaling throughout most of the length of
the veins against a predominately light background, both the distal and basal median lines
are absent, posterior valve process projects primarily inward (Figures 199 & 200) 21
20. Ventral hindwing with a thin area of black scaling (if any) throughout most of the
length of the veins or with a predominately black background, at least one median line is
present or the position of the wing where the median lines occur is solid black, posterior
valve process projects dorsally or dorsally inward (Figures 195-198 & 201-210) 28
21. Postmedian band of light colored patches or dots cream colored, ventral hindwing
discal cell at least partly connected by a thin black line (m-shaped when complete) but
lacking both the proximal and distal bands of the central symmetry system, ventral
hindwing postmedian band patches the same as the background color (cream)
Chlosyne leanira 22
21. Postmedian band of light colored patches or dots orange, ventral hindwing discal
cell lacking a thin black line but with the proximal and distal bands of the central
symmetry system present (with orange in-between except in some specimens of C. ezra),
ventral hindwing postmedian band patches (orange) contrast with the background color
(cream) 25


777
PLATE J


501
between the bands of the distal and basal symmetry system are sometimes partly fused
together. Symmetry systems bands are complete in all cells and black with orange
between them. Ventral hindwing background color between the proximal band of the
central symmetry system and distal band of the basal symmetry system cream and
concolorous with the rest of the background color (at most a very subtly lighter color).
Ventral hindwing without heavy black scaling along the veins but with thin black scaling
along parts of veins.
Geographic Variation: Chlosyne harrisii is highly variable in the amount of black scaling
on the dorsal wing surfaces. In some specimens the orange patches of the postmedian
and median bands are large and fused together with only a small amount of black scaling
between them, while at the other extreme these patches are reduced with a wide area of
black between the postmedian and median areas and between patches of each individual
band. In some specimens the orange forewing postmedian band is almost completely
absent with the cream colored postmedian dots contrasting with the black background.
Also, in lighter individuals the black markings on the forewing have an increased
diffusion of orange scales. In darker individuals the median bands may appear lighter
orange or even more cream colored than orange, although this also may appear to vary
due to differing degrees of wear between specimens. Every intermediate exists between
the darkest and lightest individuals with no gaps in the variation of any character, hence I
recognize no subspecies.
The series which I have examined that contain the lightest individuals (on the
average) are those from Manitoba. Some specimens from a variety of U. S. states are as
light as some Manitoba specimens, but the lightest Manitoba specimens are lighter than


791
Brown, F. M. 1974. The butterfly called ismeria by Boisduval and LeConte. Bull,
of the Allyn. Mus., No. 16. 12 pp.
Brown, J. W., H. G. Real and D. K. Faulkner. 1992. Butterflies of Baja
California. The Lepidoptera Research Foundation, CA, USA.
Catling, P. M, and R. A. Layberry. 1998. Distribution and biology of Chlosyne
gorgone carlotta (Nymphalidae) at its northeastern limit. J. of the Lepidopterists'
Society. 52:98-104.
Comstock, J. A. 1927. Butterflies of California. Privately published. Los Angeles,
CA.
De Quieroz, A., M. J. Donoghue, and J. Kim. 1995. Separate versus combined analysis
of phylogenetic evidence. Annual Review of Ecology and Systematics, 26: 657-
681.
dos Passos, C. F. 1938. The types of Lepidoptera described by G. D. Gunder. American
Museum Novitates. No. 999.
dos Passos, C. F. 1969. A revised synonomic list of the Nearctic Melitaeini with
taxonomic notes (Nymphalidae). J. of the Lepidopterists' Society, 23: 115-125.
Eaton, J. 1988. Lepidoptera Anatomy. Wiley, New York.
Ehrlich, P. R. 1958. The comparative morphology, phylogeny, and higher
classification of the butterflies. University of Kansas Science Bulletin, 39: 305-
370.
Emmel, J. F., T. C. Emmel and S. O. Mattoon. 1998a. New subspecies of
Nymphalidae from California and a neotype designation for Argynnis rupestris
Behr (Lepidoptera: Nymphalidae). In Systematics of Western North American
Butterflies, Thomas C. Emmel, Editor. Mariposa Press. Gainesville, FL. Pp.
139-158.
Emmel, J. F., T. C. Emmel and S. O. Mattoon. 1998b. The types of
California and Nevada butterflies named by Cajetan and Rudolph Felder:
designation of lectotypes and fixation of type localities. In Systematics of western
North American Butterflies, Thomas C. Emmel, Editor. Mariposa Press.
Gainesville, FL. pp. 95-114.
Emmel, J. F., T. C. Emmel and S. O. Mattoon. 1998c. The types of
California butterflies named by Herman Behr: designation of neotypes and
fixation of type localities. In Systematics of western North American Butterflies,
Thomas C. Emmel, Editor. Mariposa Press. Gainesville, FL. Pp. 87-94.
790


116
sides of the corpus bursae. Ductus bursae vestigial and membranous. Lamella
antevaginallis poorly developed relative to lamella postvaginallis. Lamellae lacking
ridges or extensions, and not forming a pouch enclosing the ventral genital opening.
Remarks: The phylogenetic analysis was designed to test if the Phycioditi are
monophyletic, and if so, construct a hypothesis for the placement of this clade on the
Melitaeini phylogeny. While this objective was achieved, I certainly did not examine
nearly enough Phycioditi taxa to conclude anything about the validity of the genera listed
by Higgins (1981) or other relationships within the Phycioditi, for which Higgins (1981)
included 137 species plus a number of additional subspecies. I did code some characters
which vary within the Phycioditi, and thus far, it is clear that some if not all of these
characters have been subject to reversals or independent acquisitions of similar forms.
Distribution: Phycioditi are exclusively New World (Higgins 1981). Representatives
occur throughout much North America (except in the far north), Central, and South
America.
Poladryiti Subtribe N.
Synapomorphies from binary characters: Distal posterior end of phallus with a hollow
tube open on both ends (Figs. 83,88,92 & 97-98).
Terminal derived states from multistate characters: Tegumen with paired, entire,
posterior pointed projections on the posterior lateral sides (Figs. 80,86,90 & 95) (this is a
homoplastic character which is hypothesized to have independently arisen in some
Mellicta from the most parsimonious tree). Posterior dorsal side of phallus extends distal
to the ventral posterior side as an elongate tapering extension with the sides slightly


118
Phycioditi. Vesica otherwise simple, lacking teeth disjunct from plates or granulse
patches.
Females: Ventral plate on corpus bursae variable for the presence of inverted teeth and
paired posterior extensions. Well over half of the length of the corpus bursae is free of
the ventral plate. Inverted teeth on the corpus bursae not distinctly sclerotized,
concolorous with the corpus bursae or at most very lightly sclerotized, and arranged in
either disjunct patches on the lateral sides of the corpus bursae or encircling the corpus
bursae in an irregular band widest on the sides. Ductus bursae greatly reduced and
membranous. Ostium bursae absent. Lamella postvaginallis and antevaginallis well
developed. Lamella postvaginallis with a lightly sclerotized area around the genital
opening.
Poladryas Bauer
Synapomorphies from binary characters: None.
Terminal derived states of multistate characters: Inner valve process entire with a small
tooth present just proximal to the distal end on the posterior side (Fig. 94). Posterior edge
of valve with a flattened triangular projection (Fig. 93). Surface of juxta smoothly curved
anteriorly but in approximately the posterior 1 /4th roof-shaped with a ventral keel (Fig.
94). Tegumen projections about as long as the length of the tegumen anterior to the
projection, with the outer side convex, the inner side nearly straight (slightly concave),
and the tip sharply pointed (Fig. 95). Ventral sclerotization pattern on the phallus
includes narrow inward anterior slanting extensions of sclerotized tissue in addition to
being sclerotized on the sides (Fig. 97).


787
PLATE T
m
f
m
m
IT
S:
1
Hr
Wm
18
IP
f 1
- %
19
20
i
|
' '! Yjj[ m§
4 ^mjp
*OWL
-
V1" 13
***? "tH**
14


336
model are identical to the differences between the respective boot strap consensus trees
for equal character weighting. For the remaining clades, some boot strap scores
increased, some decreased, while many remained about the same (Figure 323).
The heuristic search with the DPCWH model for polymorphisms yielded four
equally parsimonious trees prior to successive weighting, with a consistency index of
0.713, retention index of 0.888, and rescaled consistency index of 0.633. The topology of
the strict consensus tree (Figure 319) is nearly identical to the one obtained from the
other two analyses, with a few exceptions. The arrangement of the two clades of
Charidryas follows the same topology as for the distinct state model. The arrangement
of taxa within the T. leanira group is resolved with a topology that is incongruent with
both the Distinct state model and PAUPs treatment of polymorphisms. The arrangement
of taxa within the T. theona group is like that obtained from the distinct state model. C.
rosita rosita comes out as the basal taxon to a four species clade as in the analysis based
on PAUPs treatment of polymorphisms. Finally, C. marina and C. melitaeoides come
out as sister taxa, as opposed to in a trichotomy for the other two analyses. All other
clades are identical for PAUPs treatment of polymorphisms, the distinct state model, and
the DPCWH model.
After the second iteration of successive weighting (when the shortest tree was
found), the same four equally parsimonious trees and strict consensus tree (Figure 320)
were obtained. The tree statistics increased to a consistency index of 0.779, a retention
index of 0.913, and a rescaled consistency index of 0.712.
DPCWH boot strap 50% consensus trees for before and after successive
weighting are presented in Figures 321 and 322, respectively. Most of the clades


177
Figures 170-173. Male genitalia capsules of Chlosyne sp. in dorsal aspect. 170:
C. ehrenbergii. 171: C. hippodrome. 172: C. narva. 173: C. gaudealis.


768
PLATE B


486
Range: I have examined specimens from southern California, southern New Mexico,
southern Arizona, and southeast Texas as well as specimens from the Mexican states of
Sinaloa and Nuevo Leon. Collection data from the FSCA are presented in Table 12.
Stanford and Opler (1993) also report records from Sonora and Coahuila, and Brown et
al. (1992) report records from Baja California Norte and Baja California Sur. Scott's
(1986) range map includes all of Mexico north of the latitude of the southern tip of
Texas.
Species Delimitation: M. clymas has unique male and female genitalia, a unique wing
pattern, and no sister species.
Microtia coracara (Dyar)
Figures 112-113,120-121 & 279.
Phyciodes coracara Dyar, 1912. Proc. U.S. Nat. Mus. 42:40. Type Locality: Guerrero,
Mexico. Holotype: NMNH [examined],
=Phyciodes albipunctata Rober, 1914. Seitz 5:444. Type Locality: Mexico.
Diagnosis: The only certain character which I found to distinguish this taxon from its
sister taxon, M. anomalus, is the absence of a dorsal hindwing orange median band.
Also, the male genitalia of the sole representative of M. anomalus examined did not have
the saccus invaginations exposed in contrast to M. coracara-, however, I would not
conclude this is a consistent difference unless this observation can be duplicated with
additional representatives of M. anomalus. Other characters which I was able to score for


409
Bootstrap
99
60
64
78
72
78
84
62
82
93
81
74
87
85
53
84
61
84
70
98
55
74
62
81
90
87
87
96
70
64
100
66
100
62
76
75
60
97
C. nycteis
C. gorgone
C. harrissii
C. n. sp. nr. har.
C. hoffmanni
C. palla
C. gabbii
C. acastus
C. w. damoetas
C. w. whitneyi
C. definita
T. ezra
T. theona
T. perlula
T. chinatiensis
T. I. leanira
T. I. fulvia
T.cyneas
T. cynisca
C. e. pardelina
C. e. endeis
C. melitaeoides
C. marina
C. e. poecile
C. e. erodyle
C. melanarge
C. eumeda
C. californica
C. lacinia
C. ehrenbergi
C. hippodrome
C. narva
C. g. gaudealis
C. g. wellingi
C. janais
C. marianna
C. rosita rosita
C. montana
C. rosita browni
C. riobalensis
C. mazarum
T. eleda
T. coracara
T. anomalus
D. dymas
M. elva
A. pelops
A. proclea
P. minuta
H. fasciatus
H. miriam
A. perezi
A. pantoni
A. tulita
CumPhyGnaMel
Chlosyne
Microtia
Antillea
Poladryiti
Figure 314: Boot strap 50% consensus tree for successively weighted characters from the Chlosyniti/
Poladryiti data matrix, with multistate taxa interpreted as polymorphisms by PAUP.


349
pattern shared by taxa Miller and Brown (1981) placed in Charidryas, however these
similarities are symplesiomorphies. In summary, making decisions of generic limits of
Charidryas that will likely be stable over time is not possible based on currently available
evidence. The logical choice is to place C. nycteis and other taxa which have been
included in Charidryas in Chlosyne, as has already been done in several major works on
North American butterflies, including Scott (1986) and Stanford and Opler (1993). The
alternative would be to place the four taxa in the C. nycteis clade in Charidryas and errect
an additional genus for the C. hoffmani clade, but I reject this alternative because it would
require more name changes and greater deviation from classifications already used in
major works on North American butterflies.
Thessalia Scudder
The phylogenetic analyses clearly indicate that Thessalia Scudder is a well
supported monophyletic genus. The problem is that this genus is but one lineage within
the genus Chlosyne. Retaining this genus would require erecting a monotypic genus for
C. definita, and in lieu of the above discussion of Charidryas, a new concept of
Charidryas limited to the four taxa in the clade with C. nycteis, and another genus for the
clade including C. palla. Consequently, the fewest name changes required to create a
natural classification scheme involves transferring the taxa in Thessalia to Chlosyne.
Anemaca Kirby
Higgins (1960) rejected placing ehrenbergi in a separate genus because the
structural differences do not seem sufficiently important to justify separating ehrenbergi
into another genus (Higgins 1960). I reject the placement of ehrenbergi in the


306
#2=Under wing surface dots distinctly smaller and different in appearance,
composed of brown (not black) scales.
3=Absent on the dorsal surface (except for one apical dot), present on the ventral
surface.
Taxa lacking postmedian dots are designated *=?, since some out group taxa have
postmedian dots and others do not.
112. For those taxa with character state 1 of the second preceding character, the
occurrence of cream colored scales within the postmedian dots:
l=Often present, but absent in a portion of individuals.
2=Present in all individuals.
State 1 taxa were not coded as a polymorphism because individuals of state 1 taxa exhibit
one range of variation, while individuals of state 2 taxa exhibit a different range of
variation and these ranges of variation do not overlap. Cream colored scales vary from
none present to only a few present in one dot to many present in state 1 individuals.
Flowever, individuals of state 2 taxa always have distinctly more cream colored scales
and fewer black scales compared to any state 1 individual. Taxa lacking state 1 of the
second preceding character are coded ?.
113. Given state 1 of the third preceding character, the extent of cream colored scaling
within the hindwing postmedian dots on the dorsal wing surface.
l=Cell M3 dot sometimes with a small amount of cream colored scales in the
center, but sometimes with none. Other postmedian dots lack cream colored
scales.


348
Charidryas Scudder has been applied by some authors (Ferris and Fisher 1977,
Miller and Brown 1981, Ferris 1989) to nycteis, gorgone, harrissii, hoffmani and the
palla group, while the genus has not been recognized by others such as Higgins (1960)
and Scott (1986). The type species for Charidryas is C. nycteis (Higgins 1960). Higgins
(1960) rejected the genus in his revision because The specialized characters do not
appear to be sufficiently marked to justify generic separation (Higgins 1960) while
Ferris (1989) felt "the species associated with Charidryas are sufficiently distinct in
maculation to merit recognition", but the issues of monophyly and stability were not
addressed.
The phylogenetic analyses provide considerable evidence that C. nycteis, C.
gorgone, C. kendallorum, and C. harrissii form a monophyletic group. However, they
were inconclusive as to whether the clade containing C. palla forms a sister clade to this
group, or a sister clade to the rest of Chlosyne, and any topology derived from an analysis
where resolution was obtained breaks down on a tree one step longer. The only character
unambiguously suggesting Charidryas may be monophyletic is character 116, the form of
the median lines, which requires one extra step on a tree where the C. pedia clade is
adjacent to the C. nycteis clade and sister to the remainder of Chlosyne. A similarity
unique to Chlosyniti and Poladryiti shared by taxa Miller and Brown (1981) placed in
Charidryas is the presence of hindwing postmedian dots composed of both black and
cream scales. However, there is no evidence that this is a terminal derived state (the two
competing topologies where it is and is not are equally parsimonious for this character-
this character state may have occurred in the ancestor that gave rise to Chlosyne and
actually be a symplesiomorphy). There are numerous additional similarities in wing


448
and the distal sides concave to straight, lateral sides of thorax with pale orange hairs
present or absent 14
11. A narrow ring of light colored scales (of a different color (orange or cream) from the
brown colored patches in the postmedian band) encircling the ventral hindwing
postmedian dots (most prominent in cell M3 and adjacent cells), basal and distal elements
of the basal and central symmetry system brown, ventral hindwing band between
elements of the central symmetry system brown (sometimes with a variable amount of
orange scaling mixed in), area between the distal band of the basal symmetry system and
proximal band of the central symmetry system a bright silvery white contrasting with the
background color 12
11 No narrow ring of light colored scales encircling the ventral hindwing postmedian
dots, basal and distal elements of the basal and central symmetry system black, ventral
hindwing band between elements of the central symmetry system orange with no brown
scaling, area between the distal band of the basal symmetry system and proximal band of
the central symmetry system concolorous with the background color 13
12. Sections of the distal median line smoothly curved and pointing distally, ventral
hindwing with a narrow orange band between the submarginal band and edge of the
wing, ventral hindwing submarginal band sections fairly straight and at most shallowly
pronounced into arrow-head or U-shaped lines Chlosyne nycteis
12. Sections of the distal median line arrow head shaped and pointing basally, ventral
hindwing lacking a narrow orange band between the submarginal band and edge of the
wing, ventral hindwing submarginal band sections pronounced into deep arrow-head or
U-shaped lines Chlosyne gorgone


686
are black or black-brown with no postmedian dots or median band, a feature unique to C.
ehrenbergi among the Chlosyniti. Ventrally the pattern of the lightest specimens has
some resemblance to C. leanira, but there is no trace of a postmedian band.
The male genitalia of C. ehrenbergi are unique, but share several synapomorphies
with C. hippodrome, C. gaudealis, and C. narva including a rounded off posterior valve
process and the presence of setae all the way to the distal end of this process (Figures
201-204). C. ehrenbergi is the only one of these taxa where the inner valve process is
convex in ventral view (Figure 150), and the process is distinctly curved but not so strong
as in C. hippodrome (Figure 151) but much stronger than in C. narva (Figure 152) and C.
gaudealis (Figure 153). The sides of the saccus fork are much farther apart in C.
hippodrome (Figure 151) than in C. ehrenbergi (Figure 150). Also, C. ehrenbergi has the
flattened posterior phallus extension curving to a blunt point (Figure 227) versus broadly
squared off in C. hippodrome (Figure 232), C. narva (Figure 233), and C. gaudealis
(Figure 229). Higgins (1960) erroneously reported that C. ehrenbergi lacked a posterior
valve process.
Further Description: Black and white but no orange scales and hairs on the labial palpi,
with the exception of only one specimen found with a few individual orange scales.
Dorsally palpi with black scales and hairs only; outer lateral side white ventrally-center
and at base while black dorsally and at tip, black hairs around the edge with some white
ones mixed in on the ventral side; ventrally with base white and black hairs and scales in
the center and throughout the terminal segment, with the white lateral bands extending
down along the sides; inner lateral side with white scales and hairs (some black may be
mixed in) dorsally and at base and black scales and hairs ventrally and at tip (some


17
have a scolus on A9, but because it occurs in a different position (dorsal to the filiform
setae), he presumed that it was not homologous to that of the Nymphalinae.
Harvey (1991) reported that systematic relationships within the Nymphalinae
were poorly understood, particularly at the generic and tribal levels. The relationships
beamong Nymphalini, Kallimini, and Melitaeini are unknown. Nymphalini and
Melitaeini share a feature unique to the Nymphalidae of having filiform setae present on
A 1,2, while Kallimini and Melitaeini share a feature unique to the Melitaeini of having
the filiform setae on A9 occurring on the sclerotized base of the scolus (Harvey 1991).
Consequently, at least one of these characters must have been subject either to reversal,
or potentially less likely, to independent acquisition.
As noted by Harvey (1991), Higgins (1981) never provided a synapomorphy for
the Melitaeini. Harvey (1991) reported the notched saccus in the male genitalia as a
unique character for Melitaeini among the Nymphalidae, but actually several Phycioditi
lack a notched saccus, so this character does not represent a universal synapomorphy.
Harvey (1991) also noted that the distribution of filiform setae on Melitaeinine larvae is
diagnostic within the Nymphalidae, based on the combination of filiform setae on Al,2
and on A9 at the sclerotized base of the scolus. However, although the combination of
these features is a good identification character, it provides no evidence of monophyly for
the Melitaeini, because each is present in some out group taxa.
The following is a summary of Higgins' (1981) higher classification of the
Melitaeini, which includes three subtribes (ranked as tribes by Higgins (1981)) and 31
genera. The species that Higgins (1981) included in each genus appear on pages 165-171
of his 1981 review of the classification of the Melitaeini.


651
yellow and concolorous with the rest of the background color. Ventral hindwing with at
most sparse black scaling along the veins where they transverse light background color.
Geographic Variation: Two distinct allopatric subspecies exist, neither of which appears
to exhibit geographic variation.
Species/Subspecies Delimitation: Chlosyne erodyle and C. poecile are included as one
species because they are allopatric, phylogenetic evidence indicates they are sister taxa,
and they differ only by a few minor but consistent pattern characters. As a single unit,
this taxon has gaps in pattern characters with all other Chlosyne. The sister taxon of C.
erodyle, C. melanarge, is sympatric and has marked gaps in pattern characters compared
with C. erodyle.
Chlosyne erodyle erodyle
Figures 164,187,223,244 & 256a-e.
Diagnosis: C. erodyle erodyle differs from C. e. poecile by having all dorsal forewing
markings basal to the postmedian dots white instead of yellow. Ventrally some of the
basal most markings (in the basal part of the discal cell and cell CuA2) may be yellow as
in C. e. poecile, but most remain white. These markings tend to be smaller in C. e.
erodyle than in C. e. poecile, but this is not a consistent character. C. e. erodyle has only
orange scales on the tarsi of the meso- and metathoracic legs, while C. e. poecile has a
mix of orange and black scales. Both sexes of C. e. poecile have the thorax ventral to the
wings covered with black hairs, and males of C. e. erodyle are like C. e. poecile but
females have white hairs mixed in with the black.


253
5=Black scales and hairs in center and throughout terminal segment but white at
base (vs. state 3), white lateral bands extend to form a white edge on both sides of
the ventral palpi.
6=Orange hairs and scales throughout except white at the base and on edges from
extensions of the white lateral bands.
7=Orange hairs (pale orange to dark orange) and scales with a thin, sparse row of
black hairs basally and on inner ventral side, variable amount of white extending
distally to a intraspecifically variable degree.
*8=Orange hairs and scales basally, distal segment and distal part of penultimate
segment with exclusively black scales and hairs.
#9=Predominately orange scales with some scattered white scales (the white
scales are largely concealed under the orange in fresh specimens).
*#A=Sexually dimorphic: female outer side predominately white with scattered
black scales and hairs, inner side predominately black scales with scattered orange
scales and dense orange hairs, white scales and hairs with some black basally;
male with predominately orange scales and orange and black hairs with white
scales mixed in with the orange, white scales and hairs basally.
@*B=2&7: Some individuals are state 2 while others are state 7.
58. Inner lateral surface of labial palpi.
0=Black scales and hairs dorsally, mosaic of black and orange scales and hairs
ventrally.


39
5=Posterior side of valve with multiple hollow processes, with the most
prominent one at the posterior ventral comer, and the next most prominent at the
posterior dorsal comer (Fig. 45).
Some Nymphalini and Kallimini with valvae very uncharacteristic of Melitaeini have
posterior valvae projections which appear highly dissimilar to any of those found in
Melitaeini. Such taxa are coded as C for this character.
19. A broad area of the posterior end of the valvae, including the area with dense hair
like setae and the barer area ventral to it, expanded into a broad posterior valve
projection:
0=None.
1=A broad dorsally curved projection with a convex ventral surface and a concave
dorsal surface (much shorter than the dorsal surface), bearing two prominent
ventral projections (Fig. 42).
20. Inner wall of ventral part of valvae turned roughly 90 and lobed, forming a pocket
shaped roughly like a half circle with many spine-like setae on the posterior surface.
l=Absent.
2=Present (Figs. 2,3,5 & 15).
The shape and lobing of this structure appears somewhat different in Eurodryas
desfontaini from the other taxa with this structure; however, a structure with identical
spine-like setae is present on the same part of the valve.
I considered coding this variation as two characters, one for the spine-like setae
and another for the lobe of the valvae; however, since the spines occur on the lobe and all


234
18. For those taxa with a hollow posterior valve process, the distal end of this process
(best seen in posterior view):
l=Sharply pointed (Figures 115,117,119,121,195-200,205-210).
2=Blunt (Figures 201-204).
Taxa lacking a hollow posterior valve process are coded 0. While H. fasciatus has a
pointed posterior valve process (dissimilar to that found in any other taxon), in order to
avoid weighting a likely independent acquisition twice H .fasciatus is coded ?, since
there is no independent evidence from other characters to suggest its hollow posterior
valve process is homologous to that of the other taxa.
19. For those taxa with a hollow posterior valve process, the distribution of setae on this
process:
l=Proximally the process contains setae, but the distal most end is bare (Figures
115,117,119,121,195-200 & 205-210).
2=The process bares setae throughout its length (Figures 201-204).
*#3=The process bare except at the very base (Figure 85).
H. fasciatus has an autapomorphic state for this character, so the potential independent
acquisition noted above is not weighted extra from this character.
20. For those taxa with a posterior valve process, the characteristics of the tapering and
bulging of this process across its length (best seen in posterior view):
1=Tapering throughout on both ends (Figures 115,117,119,121,195-199 & 208-
209).
2=Bulged basally on the ventral side, distal to the bulge tapering throughout its
length (Figure 200).


728
postmedian dots or along the costa distal to the ventral basal costal cell patch. The
ventral basal costal cell patch has more extensive orange on the anterior side in some
specimens than I have seen for subspecies montana. Neither sex has particularly elongate
forewings, and the forewing apex is at most very slightly more produced in males relative
to females.
The dorsal hindwing is solid black except for the white postmedian dots, with the
median band confined to the ventral surface. The ventral hindwing basal area is almost
solid black, with only a small amount of diffuse cream scaling. The size of the ventral
hindwing median band is variable, and a single specimen was examined which had no
median band at all. This specimen had the same associated data as other specimens with
the ventral hindwing median band normal (Mexico: Morelos, Rancho Viejo, 4 Oct. 1977,
D & J Jenkins). In all other specimens examined, the ventral median band occupied the
range of variation present in subspecies montana.
Distribution of Specimens Examined: I have seen specimens (all in the AM) only from
the Mexican state of Morelos (Rancho Viejo, Canon de Lobos, and 5 miles S of
Amacuzae). Complete data for these records is presented in Table 12. Bauer (1961) also
gives records for Alpuyeca (also state of Morelos) and two Guerrero localities (Mexcala
& Milpillas)
Chlosyne rosita mazarum Miller and Rotger
Chlosyne mazaran Miller and Rotger, 1979. Bui. Allyn Museum. 0(54): 1-4. Type
Locality and Data: Mexico: Morelos, R. de la Maza, Paraje Rancho Viejo, 26
Aug. 1966. Holotype: Allyn Museum [Examined],


605
instructions...complete synonymical catologue of Macrolepidoptera with full
bibliography.
=Melitaea alma Strecker, 1878. Lepidoptera, Rhopaloceres and Heteroceres. Part 15:
135. Lectotype Locality: Arizona, restricted to northwestern Mohave County
(Austin and Smith 1998b). Lectotype: Chicago Field Museum of Natural History
(Austin and Smith 1998b).
=Melitaea wrightii W. H. Edwards, 1886. Canadian Entomologist. 18:64. Type
Locality: San Bemadino, California. Holotype: Carnegie Museum (Higgins
1960).
=Melitaea cerrita Wright, 1905. Butterflies of the West Coast of the U.S.: 161. Type
Locality: Southern California. Lectotype: California Academy of Sciences
(Tilden 1975).
=Melitaea leona Wright, 1905. Butterflies of the West Coast of the U. S.: 160. Type
Locality: San Rafael, Marin County, California. Holotype: California Academy
of Sciences (Miller and Brown 1981).
=Melitaea wrightii form carolynae Gunder, 1926. Ent. News. 37:3. Type Locality:
Mint Canyon, California. Holotype: American Museum of Natural History (dos
Passos, 1938).
=Lemonias alma koebeli Gunder, 1927. Ent. News. 38:136. Type Locality: Inyo
Mountains, Argus Co., California. Holotype: California Academy of Sciences
(Austin and Smith 1998b).
=Melitaea wrightii form pelona Gunder, 1930. Bull. S. Calif. Acad. Sci. 29. Type


66
interpreted in one taxon, G. sodialis). Taxa with a proximal plate but lacking the second
teeth bearing plate are coded ?, while taxa lacking either plate are coded 0.
73. For those taxa with state 2 of the preceding character, the arrangement of teeth on the
teeth-bearing plate attached to the vesica:
l=Teeth in one row along an arch (Figs. 105,125 & 211-212).
2=Teeth are not confined to one row (Figs. 52,55-56,70,78,83-84 & 97).
74. For those taxa with a plate bearing teeth attached to the vesica, characteristics of this
plate (these descriptions are based on a view with the vesica everted):
1=A narrow, heavily sclerotized crescent curving posteriorly and bearing teeth
(usually 6-10) projecting posteriorly along the crescent (ridges of the anterior base
of the teeth occur on the outer side of the crescent) (Fig. 211).
2=As in state 1 except dorsally the plate extends as an anteriorly curving arc (this
state occurs in only 3 taxa of the genus Chlosyne; all other Chlosyne have state 1)
(Fig. 212).
3=Base rectangular, on posterior end a ventrally projecting, anteriorly curved, and
heavily sclerotized crescent bearing teeth as in state 1 (Fig. 105).
4=Elongate and more lightly sclerotized relative to the basal plate with an
outward arch at the very base, covered with many minute teeth on surface and
edges (Figs. 52 & 55-56).
5=Narrow crescent with prominent teeth on its surface, six teeth long anterior to
posterior and no more than three teeth wide (Fig. 70).


767
PLATE A


163
Figures 94-97. Male genitalia of Poladryas minuta. 94: Capsule (ventral). 95: Capsule (dorsal).
96: Capsule (lateral). 97: Phallus (ventral). 98: Atlantea pantoni phallus (ventral).
50X
14(1)
15(1)
16(1)
14(1),15(1)
16(


18
Higgins' (1981) higher classification of the Melitaeini (with tribes down-ranked to
subtribes):
Euphydryiti Higgins: Euphydryas Scudder, Hypodryas Higgins, Occidry as Higgins, and
Eurodrycis Higgins
Melitaeiti Tutt: Mellicta Billberg, Melitaea Fabricius, Poladryas Bauer, Didymaeformia
Verity, Cinclidia Hubner, Chlosyne Butler, Thessalia Scudder, Texola Higgins, Dymasia
Higgins, Microtia Bates, Gnathotriche Felder & Felder, Gnathotrusia Higgins,
Higginsins Hemming, Antillea Higgins
Phycioditi Higgins: Phyciodes Hubner, Phystis Higgins, Anthanassa Scudder, Dagon
Higgins, Telenassa Higgins, Ortilia Higgins, Tisona Higgins, Eresia Boisduval, Castilia
Higgins, Janatella Higgins, Mazia Higgins
Unplaced Genus: Atlantea Higgins
The support of a final phylogenetic tree can be no stronger than the support of any
assumption made along the way, and because phylogenetic trees generated in parsimony
analysis with morphological data are rooted and the characters are then polarized based
on the out group (Kitching 1992a), the out group decision is critical. Although the best
evidence currently available suggests that Nymphalinae forms a monophyletic group
including Nymphalini, Kallimini, and Melitaeini (Harvey 1991), the relationships among
these three tribes are unclear (Harvey 1991). Consequently the closest relatives of the
Melitaeini are unknown. A similar problem of out group uncertainty was addressed with
respect to a phylogenetic study of Rekoa Kaye (Lycaenidae: Theclinae) by Robbins
(1991). Robbins (1991) presented evidence that Rekoa was related to Arawacus Kaye
and Thereus Johnson, but was unsure which represented the sister group and unsure of
the of relationships within these two genera. Robbins (1991) solved this problem by
including taxa in these two genera as the out group taxon, and by coding them


525
where the femur tucks in with black scales and hairs and some white scales underneath,
thorax ventral to the wings densely covered with white and pale orange scales and hairs
with some dark scales under the white ones. Antennal shaft with a checkered pattern of
black and white scales with or without orange scales, club with a white patch on the outer
lateral side.
Narrow band of orange scales along the base of the forewing costal margin. Wing
fringes checkered black and white with white sections between where the veins reach the
wing margin. Basal part of ventral forewing cell C orange along the edges of the cell and
predominately white in the middle. Discal spot present with orange scaling inside.
Forewing discal cell with a patch formed from the fusion of the distal band of the basal
symmetry system and proximal band of the central symmetry system with orange scaling
inside the same color orange as that inside the discal cell. Dorsal forewing discal cell
with a single orange to cream colored patch between the proximal band of the central
symmetry system and the discal spot. Basal band of basal symmetry system forming a
basal patch in the discal cell with diffuse orange scaling of identical color to the orange
inside the discal spot. Dorsal forewing cell CuA2 with the distal element of the basal
symmetry system single and forming a narrow, elongate patch (with diffuse orange
scaling inside) extending from the basal part of the cell almost to the origin of vein
CuA2. Dorsal hindwing postmedian band present and in the form of a series of orange
patches in cells CuA2-R5, ventrally patches in cells CuA2-Sc+Rl with orange patches in
CuA2-Ml and cream background color and a variable amount of black in cells R5-Sc+Rl
(Sc+Rl may contain a black patch). Forewing postmedian band of orange patches with
cream colored postmedian dots inside. Hindwing postmedian dots present within the


393
Table 5
Continued
Characters
Taxa
91 92 93 94 95
96 97 98 99
too
101
102 j 103 104
105
106
107
108
109
110
111
112
113
114
115
116 I 117
118
119 120
Chlosyne nycteis
0 0
0
0
0
0 1 *
0
0
0
0 0?
1
2
0
0
1
1
1
1
2
1
0
2 0
1
1 1
Chlosyne gorgone
0 0
0
0
0
0 1 *
0
0
0
0 0?
1
2
0
0
1
1
1
1
1
1
0
3 0
5
* 4
Chlosyne harrissii
0 0
0
0
0
0 1 *
0
0
0
0 0?
0
1
0
0
1
1
1
1
2
0
0
4 0
1
1 1
Chlosyne kendallorum
0 ?
0
0
0
0 1 *
0
0
0
0 0?
0
1
0
0
1
1
1
1
2
0
0
4 0
1
1 1
Chlosyne hoffmanni
0,2 0
0
0
0
0 1 *
0
0
0
0 0?
0
1
0
0
1
1
1
2
3
0
0
4 1
1
1 1
Chlosyne palla
0.2 0
0
0
0
0 1 *
0
0
0
0 0?
0
1
0
0
1
1
1
2
3
0
0
4 1
1
1 1
Chlosyne gabbii
0.2 0
0
0
0
0 1 *
0
0
0
0 0?
0
1
0
0
1
1
1
2
3
0
0
4 1
1
1 1
Chlosyne acastus
0.2 0
0
0
0
0 1 *
0
0
0
0 0?
0
1
0
0
1
1
1
2
3
0
0
4 1
1 1
Chlosyne whitneyi damoetas
0,2 0
0
0
0
0 1 *
0
0
0
0 0?
0
1
0
0
1
1
1
2
3
0
0
4 1
1
1 1
Chlosyne whitneyi whitneyi
? 0
0
0
0
0 1 *
0
0
0
0 0 ?
0
1
0
0
1
1
1
2
3
0
0
4 1
1
1 1
Chlosyne definita
2 0,1
0
0
0
0 1 *
0
0
0,3
0 0?
0
1.3
0
0
1
3
1
0
0
*
0
0 2
0
* 0
Thessalia ezra
2 0
?
1
0
0 1 *
0
0
?
0 2? 2
*
4
1.2
0
1
0
'
0
0
*
1
* 2
0
* 0
Thessalia theona
2 0
2
0.1
0
0 1 *
0
0
2
0 2 2
*
4
1.2
0
1
0.6
*
0
0
*
1
* 2
0
* 0
Thessalia perlula
2 0
2
0.1
0
0 1 *
0
0
2
0 2 2
*
4
1.2
0
1
0.6
?
0
0
*
1
* 2
0
* 0
Thessalia chinatiensis
2 0
2
1
0
0 1
0
0
2
0 2 2
*
4
1.2
0
1
0
*
0
0
*
1
* 2
0
* 0
Thessalia leanira leanira
2 0
2.1
1
0
1 1 *
0
1
0
0 2 1
*
5
0
0
1
0
*
0
0
*
1
* 2
0
* 0
Thessalia leanira fulvia
2 0
2.1
1
0
1 1 *
0
1
0
0 2 1
*
5
0
0
1
0
*
0
0
*
1
* 2
0
' 0
Thessalia cyneas
2 0
?
?
0
1 ? *
0
1
0
0 2 1
*
5
0
0
?
0
*
0
0
*
1
* 2
0
* 0
Thessalia cynisca
2 0
?
?
0
1 ? *
?
*
*
0 2? 1
*
5
0
0
1
?
*
0
0
*
1
* 2
0
* 0
Chlosyne endeis pardelina
1 0.1
0.2
0
0
0 1 *
0
0
4
2 0.2 2
0
1.3
0
1
1
3
1
0
0
*
0
0 2
0
* 0
Chlosyne endeis endeis
N ?
0.2
?
0
0 1 *
0
0
4
2 0.2 2
0
1.3
0
1
?
3
1
0
0
*
0
0 2
0
* 0
Chlosyne melitaeioides
1 0.1
0
0
0
0 ? *
0
0
1.4
2 0,2 2
0
1.3
0
1
1
3.4
1
0
0
*
0.2
? 2
0
* 0
Chlosyne marina
1 ; 0.1
0
0
0
0 1 *
0
0
1.4
2 0,2 2
0
1.3
0
1
1
3.4
1
0
0
*
0.2
0 2
0
* 0
Chlosyne poecile
3 o.t
0
1
0
o ? *
0
0
5.6
1 3 ?
*
3
0
1
?
4
1
0
0
0
2
0 2
0
0
Chlosyne erodyle
1 0,1
0
1
0
0 ? *
0
0
5.6
1 3 ?
3
0
1
?
4
1
0
0
0
2
0 2
0
* 0
Chlosyne melanarge
' *
?
?
0
0 ? *
0 0
6
1 3 ?
*
3
1
*
?
o,:
*
0
0
*
?
? 2
0
* 0
Chlosyne hylaeus
0
?
0
0 ? *
0
0
1
2 3?
*
1
0
0
?
3
1
0
0
0
2
0,1 2
0
* 0
Chlosyne eumeda
2,3 0.1
0
?
0
0 ? *
0
0
1
2 3?
*
1
0
0
?
0
1
0
0
*
2
1 2
0
* 0
Chlosyne californica
2 0
?
?
0
0 ? *
1
*
*
0 4?
*
*
0
1
?
2
1
0
0
0
?
* 2
0
* 0
Chlosyne lacinia
1,2 0
0
?
0
0 1 *
1
*
*
0 4?
*
*
0
1
?
2
1
0
0
0
1
* 2
0
* 0
Chlosyne ehrenbergii
* *
1
r
0
1 2 *
2
*
4 ?
*
*
1
*
1
0
*
0
0
*
1
* 2
7
* 6
Chlosyne hippodrome
* *
?
0
0 ? *
1
*
*
0 4?
*
0
1
?
2
1
0
0
0
?
? 2
0
* 0
Chlosyne narva
? ?
1
*
0
0 2 *
1
*
*
0 4?
*
*
0
1
1
2
1
0
0
0
?
* 2
0
* 0
Chlosyne gaudealis
? ?
1
*
1
0 2 *
1
*
*
0 4 ?
*
0
1 1
2
1
0
0
0
?
* 2
0
* 0
C. gaudealis wellingi
* *
?
?
1
0 ? *
1
*
0 4?
*
0
1
?
2
1
0
0
0
?
? 2
0
* 0
Chlosyne janais
1 1
?
0
0
0 ? *
1
*
*
0 4?
*
0
1
1
2
1
0
0
0
2
1 2
0
* 0
C. marianna
1 1
?
?
0
o ? *
1
*
*
0 4?
*
*
0
1
?
2
1
0
0
0
?
? 2
0
* 0
Chlosyne rosita rosita
1 0
?
?
0
0 ? *
1
*
*
0 4?
*
0
1
?
2
1
0
0
0
2
1 2
0
* 0
Chlosyne montana
1 0
?
?
0
0 ? *
1
*
*
0 4?
*
*
0
1
?
2
1
0
0
0
2
1 2
0
* 0
Chlosyne rosita browni
1 0
?
?
0
0 ? *
1
*
*
0 4?
*
0
1
?
2
1
0
0
0
2
1 2
0
* 0
Chlosyne riobalensis
1 0
?
?
0
0 ? *
1
*
*
0 4?
*
*
0
1
?
2
1
0
0
0
?
? 2
0
* 0
Chlosyne mazarum
N 0
?
?
0
0 ? *
1
*
*
0 4?
*
*
0
1
?
2
1
0
0
0
?
? 2
0
* 0
Texola elada
0 0
0
0
0
0 0 0
0
0
0.3
0 0?
0
0
1
0
0
0
* I
0
0
0
0
0 2
1
1 1
Texola coracara
1 0
0
0
0
0 0 *
?
*
*
0 1 ?
*
*
0
1
0
2
1
0
0
0
0
1 2
1
2 1
Texola anomalus
? 0
0
0
0
0 ? *
?
*
*
0 1 ?
*
*
0
1
?
2
1
0
0
0
0
1 2
1
2 1
Dymasia dymas
0 0
0
0
0
0 0 0
0
0
3
0 0?
0
0
1
0
0
0
*
0
0
0
0
0 2
2
1 1
Microtia elva
? ?
?
?
?
? ? ?
?
?
?
? ? ?
?
?
1
?
?
0
?
?
?
?
?
? ?
?
? ?
Antillea pelops
0 0
4
0.1
0
0 0 0,1
0
0
0
0 5?
*
?
1.3
1
0
7
3
0
0
0
3
0 0.1
8
? 7
Antillea proclea
0 0
4
0
0
0 0 0
0
0
0
0 5?
*
?
3
2
0
7
3
0
0
0
3
0 0.1
8
? 7
Poladryas minuta
0 0
0
0
0
0 0 0
0
0
3
0 0?
0
0
0
0
0
0.6
1
0.1
0
0
0
1 0
4
* 3
Higginsius fasciatus
N 0
0
0
0
0 0 0
0
0
0
3 ? ?
?
?
?
?
?
6
2
1
0
0
0
0 3
6
* 5
Higginsius miriam
4 0
?
?
0
0 0 0
0
2
0
3 ? ?
?
?
0
0
?
6
2
1
0
0
0
0 3
6
* 5
Atlantea perezi
0 0
3
1
0
0 0 1
0
0
?
3 3 ?
*
6
0
1
0
0
*
0
0
0
?
? 2
3
* 2
Atlantea tulita
0 0
3
0
0
0 0 0
0
0
?
3 3?
*
6
0
2
0
0
*
0
0
0
0
0 2
3
* 2
Atlantea pantoni
3 0
3
1
0
0 0 1
0
0
0
3 3?
*
6
0
2
0
0
*
0
0
0
0
0 2
3
* 2
? Not coded (see character descriptions for explanation)
Indicates a taxon lacks a primitive character that is subdivided into forms as a separate character (=?)
N No material examined i
=?)


726
thin and not apparent without close inspection, which is never the case in subspecies
browni or rosita. However, one unique specimen from the Barnes collection labeled
"Arizona" has all the characteristics of subspecies montana that occur outside the range
of variation of subspecies browni except the dorsal hindwing median band is as large as
in browni specimens.
On the ventral forewing surface the elongate cream patch is present in the basal
costal cell where it is usually (but not always) bordered anteriorly by a thin row of orange
scales. In a few specimens (2/13 checked for this character including a male and a
female (more specimens were examined for the other characters)), cream scaling
continues along the costal margin as a narrower row and then widens again just basal to
the median band in the R cells, and then extends slightly distal to the R cells as a narrow
row. Some specimens lacking the row of cream along the costa have the median band
patches in cells R4 and R3 surrounded by a thin area of cream scales. Some individuals
also have cream scaling at the distal edge of some of the postmedian dots on the ventral
forewing surface, and not always confined to the apical most dots as in subspecies
browni, although some specimens have no cream edge to any postmedian dots. The
ventral forewing light markings are more conspicuous than on the dorsal surface with less
contrast in size between the anterior patches of the median band and other markings. In
contrast to subspecies browni, on the ventral hindwing the median band and light basal
area are almost always separated by an area of black of variable width, though sometimes
the black is very thin and the separation is not always complete. The extreme specimens
of browni with the most separation are near the extreme specimens of montana with the
least separation, but usually the separation in subspecies montana is as extensive as in


150
45X
Figures 37-39. Male genitalia capsule of Melitaea cinxia. 37: Ventral. 38: Dorsal.
39: Lateral.
18(4)


619
touching the eyes ventro lateral of the antennal bases and a variably sized white patch
centered ventrally on the face, with black scales and hairs and sometimes scattered white
ones elsewhere. Black and white but no orange scales present in the collar between the
head and pronotum in males, but orange scales also present in females. Posterior edges
of posterior abdominal segments with bands of white scales on dorsal and lateral sides,
black and a variable mixing of orange scales anterior to these bands. Femur of meso- and
metathoracic legs with predominately orange scales dorsally and white scales ventrally
except at distal end where only orange scales are present. Wing pattern features not
common to all C. leanira subspecies are discussed below.
Geographic Variation: There is some geographic variation within subspecies fulvia but
much less so than in subspecies leanira. The variation occurs along a continuum, hence I
do not recognize any additional taxa under the name fulvia. Smith and Brock's (1988)
range map indicates the areas they designate for the three subspecies they recognize and
populations they do not assign to one of their subspecies. I have examined specimens
from within the area designated for the three subspecies and "undetermined" populations,
but from many fewer localities, and only two paratypes from the area designated for
pariaensis.
Males do not appear to exhibit more than minimal geographic variation, other
than for average size. They are variable within populations, but a similar range of color
and pattern variation appears to occur throughout the range, including specimens with
nearly as much orange as the alma phenotype of subspecies leanira, specimens which are
boldly patterned with black, and specimens which are predominately pale cream with no
more than very sparse and diffuse dorsal orange limited to cells M3 and M2 in the


at the generic level, and includes three monophyletic genera: Atlantea Higgins,
Higginsius Hemming, and Poladrycis Bauer. Chlosyniti is revised in detail at the generic
and species levels, and includes three monophyletic genera: Antillea Higgins, Microtia
Bates, and Chlosyne Butler, with the genera Dymasia Higgins, Texola Higgins,
Charidryas Scudder, Thessalia Scudder, and Anemaca placed into synonymy. Thirty-
seven species taxa and seventeen subspecies taxa are recognized within the Chlosyniti.
Detailed camera lucida drawings illustrate all genitalic characters and character states.
The phylogenetic analysis of Poladryiti and Chlosyniti is used as a case study to
investigate some issues of broader implication to systematic biology, including equally
versus successively weighting characters, tree statistics, homoplasy, boot strap scores,
and polymorphisms. Separate analyses of genitalic and pattern characters were
conducted for the Cholsyniti/Poladryiti data matrix to investigate the proportion of
groupings in conflict between independent data sets for different methods of analysis.
The percentage of conflicting groupings was 0% for equal weighting of characters with
parsimony, 36.8% for successive weighting and parsimony, 77% for a phenetic algorithm
(UPGMA), and 100% between pairs of random trees. The absence of incongruence
between independent data sets for the former analysis is argued as evidence supporting
the effectiveness and theoretical validity of the phylogenetic methods used in this study.
x


488
(lacking orange) with white patches dorsally and on each side of the posterior edge of
abdominal segments, ventrally with off white scales and paired black longitudinal stripes.
Prothoracic tibia and tarsi with white scales and hairs throughout. Femur and tibia on the
meta- and mesothoracic legs with orange scales dorsally and white scales ventrally, tarsi
as in M. eleda (above). Antennal shaft with the characteristic black and white checkered
pattern and without orange scales, club with a outer lateral white patch.
Forewing and hindwing fringe of scales checkered black and white, with white
sections between the ends of the veins. Ventral forewing basal part of cell C with brown
and light orange scales. Discal spot present but may only be detectable on the ventral
surface (the ventral forewing background is dark but the outline of the discal cell is
darker). Patch in forewing discal cell formed from the fusion of the distal band of the
basal symmetry system and proximal band of the central symmetry system present, but
also may only be detectable ventrally. A single white colored patch present between the
discal spot and proximal band of the central symmetry system, and sometimes between
the distal and basal bands of the basal symmetry system. Basal element of the basal
symmetry system in ventral forewing discal cell forming a finger like basal patch with a
variable amount of white scaling inside. When detectable (ventrally only), the distal
band of the basal symmetry system in forewing cell CuA2 is double. Ventral hindwing
postmedian band occupying cells CuA2-Sc+Rl and formed by diffuse black scaling
between the distal and basal parafocal elements against a light background, not detectable
dorsally against a black background. Forewing and hindwing with row of white
postmedian dots. When detectable (only ventrally), forewing cell CuA2 with the basal
element of the basal symmetry system present. Distal and basal median lines present on


470
examined (Dymasia and Microtia were monotypic in Higgins' (1960 and 1981) concepts)
in addition to at least one of the other three I place in Microtia. Knowledge of the form
of this character in Antillea and primitive Chlosyne clades would be important to
evaluating its utility as a synapomorphy for Microtia.
Terminal derived states of multistate characters: Sclerotization pattern on the ventral
phallus surface with the lightly sclerotized area ovoid and the darkly sclerotized area
reduced to an extremely narrow band at one point and expanding anterior and posterior of
this point (Figures 127,129,131 & 133). Posterior valve process projecting strongly
inward and curved slightly dorsally at the tip (Figures 115,117 & 121) except in M.
dymas which has an autapomorphic state (Figure 119). Grooves for the femur in the
meso- and metapleurons with light orange and black scales and hairs present (this is a
homoplastic character which appears to have independently evolved in one Chlosyne and
in Poladryas). Dorsal hindwing postmedian band patches (when present) occupying
seven cells from CuA2 to Sc+Rl (this character state appears to have independently
evolved in Poladryas).
Further Description: Male Genitalia: Distal teeth bearing plate attached to the vesica
with the teeth arranged along a convex (in posterior to basal aspect) arch with no
posterior extension on its dorsal edge. Posterior dorsal end of phallus with a flattened
extension with a broad convex base. Anterior extent of the lightly sclerotized area on the
ventral phallus surface terminating posterior to the supersensory membrane except in
some individuals o Microtia elva. Ventral valve process entire throughout except
pointed apically. Posterior valve projection present, tapering and pointed apically, and
with the terminal area free of setae. Posterior edge of inner valve process convex in


600
is convex and the anterior side concave in ventral view with the tip projecting
inward(Figure 147) or rotated at the base approximately 90 degrees such that the tip
projects ventrally (Figure 146) or any intermediate condition. The saccus fork varies
from very shallow (Figures 146-147) to substantially deeper (Figure 169). These
characters do not vary geographically or between the different subspecies.
Smith and Brock (1988) provide a key to the four subspecies of C. leanira which I
recognize plus several additional subspecies they recognize. However, I find all of the
characters in the key (except for the first character of couplet 3) do not provide reliable
separation because the ranges of variation exhibited by the taxon or group of taxa at each
dichotomy partially or completely overlap for the characters given and/or the characters
are not accurate. For example, couplet 1 splits off cynisca with "small white submarginal
[=postmedian] spots on both wings, white median spots on forewing and a continuous
cream-yellow median band on HW". The postmedian band and forewing median band
patches are never white in any subspecies of C. leanira, and these markings vary from
pale whitish cream to yellow tinged cream in all four subspecies I recognize, including
cynisca. Many specimens of the other three subspecies have the hindwing median band
as continuous and as yellow as subspecies cynisca. Characters I find to be reliable for
separating the subspecies of C. leanira are provided in the diagnosis of each subspecies
and in key at the beginning of this chapter. It is not possible to write a key to any of the
additional subspecies under the names leanira and fnlvia recognized by some authors, as
to my knowledge there are no gaps in the continuum of variation exhibited within each of
the four subspecies which I recognize.


278
87. Presence of light colored scaling in the center of the discal spot (on the ventral or
both dorsal and ventral wing surfaces):
0=Present.
l=Absent.
@A=0&1: Varies along a continuum between state 0 and state 1.
Note that the extent of light colored scaling in the discal spot is more extensive on the
ventral surface or equally extensive on the dorsal and ventral surfaces.
Some individuals of C. marina and C. melitaeoides have diffuse orange scaling in
the discal spot ventrally, whereas in other individuals the discal spot is entirely black
ventrally. These taxa are coded 0,1 for this character.
In Nijhouts (1991) Nymphalid ground plan, he diagrams two patches in the
middle of the forewing discal cell (C & D), termed the Basal Symmetry System and the
Proximal Band of Central Symmetry System, respectively. Some taxa in the Chlosyniti
lack these two features, but many taxa have a single wider patch in the center of the
forewing discal cell, apparently derived from a fusion of these two bands in the
Nymphalid ground plan. This is well evidenced by the variation present in Polydryas
minuta and Dymasia dymas, where some specimens appear to have the two black bands
as in Nijhouts illustration, while in other specimens the bands are merged together as
characteristic of the Chlosyniti. This patch spans the height of the cell, and usually has a
black border with light scaling in the middle. The black border and light scaling exhibit
the same range of intraspecific variation as in the discal cell. The basic design of this
patch is three expanded light areas in the inside divided completely or incompletely by a


195
Figures 273-276. Female genitalia of Phycioditi and Poladryiti. 273: Tisona saladellensis
(ventro-posterior aspect). 274: Atlantea tulita (ventral). 275: Lamellae of same (lateral).
276: Higginsius miriam (ventral).


456
to the ventral hindwing postmedian patches in cells CuA2-CuAl and M2-M1 less than
the basal to distal width of these patches 37
36. Dorsal forewing cell M3 with a small cream postmedian dot and with the red-orange
postmedian band patch reduced to a thin border of scales around the postmedian dot
Chlosyne hylaeus
36. Dorsal forewing cell M3 without a postmedian dot but with a well developed red-
orange postmedian band patch Chlosyne eumeda
37. Forewing median band white or cream in both sexes (never yellow), non black scales
in dorsal hindwing basal area cream colored (may have a pale yellow tint), cell M3 never
with a complete red-orange dorsal hindwing postmedian patch and with a postmedian dot
(sparse red-orange scaling may occur around the postmedian dot) Chlosyne marina
37. Forewing median band yellow in males and white in females, non black scales in
dorsal hindwing basal area deep yellow, females (always?) with a complete red-orange
dorsal hindwing postmedian patch in cell M3 and no postmedian dot (males as above)...
Chlosyne melitaeoides
38. On the ventral hindwing cells 1A+2A and CuA2 with orange extending as far
distally as the postmedian dot in cell CuA2 39
38. On the ventral hindwing cells 1A+2A and CuA2 either lacking orange entirely or if
present not extending as far distally as the postmedian dot in cell CuA2 40
39.Ventral forewing discal cell predominately orange, thin area of orange scaling along
the ventral hindwing and forewing margin, ventral hindwing cells CuA2 and both anal
cells have orange continuous with adjacent cells and extending basally to the wing base,
dorsal forewing and hindwing always with orange (distally) and cream (basally) scales in


145
23(1),24(1)
:0(2),21(1)
39(1),40(1)
Character numbers are from the Melitaeini data matrix (Table 1)
5
Figs. 2-5. Male genitalia of Euphydryiti. 2: Eurodryas aurinia (ventral). 3: Same
(posterior inverted). 4: Same (dorsal). 5: Hypodryas gillettii (ventral).


334
clades appearing in the boot strap consensus trees are incongruent with either strict
consensus tree except for a clade of (eleda(elvci,dymas)) in the analysis with equally
weighted characters, which is not only not supported by parsimony but actually
represents a less parsimonious arrangement requiring two extra steps.
The heuristic search of equally weighted characters with the distinct state model
for polymorphisms yielded 27 equally parsimonious trees with a consistency index of
0.754, a retention index of 0.894, and a rescaled consistency index of 0.675. The strict
consensus of these trees is presented in figure 315. 139 characters were parsimony
informative, and 67 characters were homoplastic, including seven genitalic characters and
60 pattern characters.
A notable difference between this consensus tree and that obtained from PAUPs
treatment of polymorphisms is that the Chciridryas group does not come out as
monophyletic, but the clade with C. harrissii comes out basal to the clade with C. palla,
which is basal to the remaining Chlosyne. However, a topology with the Charidryas
group monophyletic requires only one additional step. There are several additional
differences in terminal taxa with respect to the preceding strict consensus tree: (1) there is
no resolution within the T. leanira group, (2) the topology of the T. theona three species
clade is different, and (3) the position of Chlosyne rosita rosita is unresolved rather than
basal to the other four taxa in the C. rosita group (browni, montana, riobalensis, and
mazarum).
After two iterations of successive weighting, the minimum tree length was
obtained and nine equally parsimonious trees were generated. These trees have a
consistency index of 0.829, a retention index of 0.925, and a rescaled consistency index


354
subspecies, given the absence of any evidence to suggest rosita and browni are most
closely related, and some evidence to suggest they are not. The presence of a red-orange
basal patch/band on the dorsal wing surface in browni, rosita, and montana may well be a
symplesiomorphy with respect to the C. rosita ancestor, since the patch is also present in
some individuals of C.janais, the sister taxon to the rosita group. Since the patch blends
from present to absent in C.janais, and is present in some C. rosita subspecies but not
others, based on this character alone it is equally parsimonious that the patch could have
been independently acquired in the rosita group versus present in the ancestor and
subsequently lost. The presence of a yellow band instead of a red orange one in C.
montana is an autapomorphy without value for hypothesizing relationships.
The five taxa in the rosita group all appear to be allopatric, are clearly
monophyletic, and all have identical genitalia. The differences separating them are only
a few minor differences in pattern characters. Regardless of the weakly supported
phylogenetic relationships between these taxa, by my subspecies concept ranking them as
five subspecies of Chlosyne rosita is the appropriate decision.
Chlosyne leanira, Chlosyne fulvia, Chlosyne cyneas, and Chlosyne cynisca
The species level classification of C. leanira, C. fulvia, C. cyneas, and C. cynisca
has varied with different publications, yet none of the species level classifications of
these taxa in the literature appear to be based on evidence of relationships. Higgins
(1960) included all of these taxa as members of Thessalia leanira. Higgins (1981) later
split cynisca out as a separate species. Scott (1986) considered C. leanira and C. fulvia to
be the same species and presented C. cyneas as a separate species but suggested it may be
another subspecies of C. leanira. Miller and Brown (1981) and Stanford and Opler


406
C. nycteis-
C. gorgone
C. n. sp. nr. har.
C. harrissii
C. hoffmanni
C. palla
C. gabbii
C. acastus
C. w. damoetas
C. w. whitneyi
C. definita
T. ezra
T. theona
T. perlula
T. chinatiensis
T. I. leanira
T. I. fulvia
T. cyneas
T. cynisca
C. e. pardelina
C. e. endeis
C. melitaeoides
C. marina
C. e. poecile
C. e. erodyle
C. melanarge
C. eumeda
C. californica
C. lacinia
C. ehrenbergi
C. hippodrome
C. narva
C. g. gaudealis
C. g. wellingi
C. janais
C. marianna
C. rosita rosita
C. montana
C. rosita browni
C. riobalensis
C. mazarum
T. eleda-
T. coracara
T. anomalus
D. dymas
M. elva
A. pelops
A. proclea
P. minuta
H. fasciatus
H. miriam
A. perezi
A. pantoni
A. tulita
Chlosyne
Microtia
Antillea
Poladryiti
CumPhyGnaMel
Figure 311: Strict consensus tree from a heuristic search of equally weighted characters from the
Chlosyniti/Poladryiti data matrix, with multistate taxa interpreted as polymorphisms by PA UP.


294
to diffuse scaling. The other taxa varied intraspecifically at most by whether one of the
patches was present, absent, or reduced to diffuse scaling.
C. hylaeus is coded state 1; however, in the two individuals examined the red-
orange scaling in cell M3 was extremely reduced (to a thin border around the postmedian
dot).
There is a small amount of overlap between state 0 and state 1. In some of the
taxa with state 0, a small proportion of specimens occasionally have the patch in R5
reduced to diffuse scaling, while in most it is prominent. For those taxa with state 1, the
patch is R5 is never prominent, usually absent, and sometimes present as diffuse scaling.
If the R5 patch could be quantified for a series of specimens for each relevant taxon, it is
obvious that there would be a wide and significant gap in average variation between but
not within state 1 and 0, even though there are occasional specimens among some state 0
taxa that are closer to state 1. Consequently, it seems more appropriate to split 0 and 1
into two states.
For the four specimens examined of the out group taxon Atlantea tulitci, two
females and one male have the usual six patches whereas one male has only three (the
CuAl through M2 patches are absent). The two female specimens of Atlantea perezi
examined lack the cell R5 patch, and the two males have continuous red-orange scaling
in cell R5 from this position and basal to it. I code both of these taxa as ? for this
character.
Taxa lacking a hindwing postmedian band are coded *=?. This is also true of taxa
with state 1 for the second preceding character, as due to the hindwing around the


179
Figs. 176-185. Male genitalia capsules of Chlosyne sp. in lateral aspect. 176: C. nycteis.
177: C. gorgone. 178: C. harrissii. 179: C. kendallorum. 180: C. hoffmani. 181: C. acastus
182: C. definite!. 183: C. endeis. 184: C. theona. 185: C. leanira.


589
large black bordered patch at the distal end and spanning the width of the discal cell, and
a smaller black bordered patch basal to the position of the distal band of the basal
symmetry system in the anterior portion of the discal cell (occasionally absent or very
minute in some of the darkest phenotypes). An additional concolorous black bordered
patch occurs between the distal band of the basal symmetry system and basal band of the
central symmetry system in cell CuA2. In specimens with the discal cell predominately
orange, these two symmetry system bands are fused together in the discal cell forming a
patch with sharp or diffuse orange scaling inside. The distal band of the central
symmetry system is often broadly expanded basally in cells M2-R3, but usually with
some orange between the distal edge of the discal cell and the black area. In some
specimens at first glance this orange may appear deceptively like an orange filled discal
spot, but close examination reveals C. theona never has a discal spot, and the vein at the
distal edge of the discal spot is lined with black scaling. The area between the central
symmetry system bands in cells CuA2-M3 varies between predominately orange with
diffuse black scaling inside to solid black.
The ventral forewing includes the same light colored markings present dorsally,
but they are often at least slightly larger. The series of patches distal to the postmedian
band includes a well defined series of patches in cells CuA2-R3, with the largest patch in
M3 followed by (not in a clear and consistent sequence of relative sizes)) CuAl, M2, R4,
and R5, and the smallest patches in cells CuA2 and Ml. The postmedian band includes a
well defined patch in cells CuA2-CuAl and M3-R3 even if this was not the case dorsally,
with the cell M3 patch variably present or absent. In light colored specimens the amount
of black scaling ventrally is similar or sometimes slightly more extensive than dorsally.


697
Forewing basal costal margin lacking the narrow band of orange scales, only
black scales present. Wing fringes checkered black and white with white sections
between where the veins reach the wing margin, and the forewing white section across
from Ml tending to be relatively small.
Basal part of ventral forewing cell C solid black without a distinct patch of
orange. Dorsal forewing and hindwing with a prominent and complete row of white
postmedian dots with solid black distal to this area. Remainder of dorsal forewing with
light markings (pale yellow, deep yellow, or with a variable amount of orange) against a
black background. Median band present in cells CuA2-CuAl and M2-R5 but absent in
M3. Dorsal forewing discal cell with two disjunct light patches, one just basal to the
position where the discal spot would occur (the discal spot is not detectable against the
black background if it is present), which is usually a single patch but occasionally divided
into a small anterior and larger posterior section by diffuse black scaling, and another
elongate patch in the basal part of the discal cell of variable width anterior to posterior,
and sometimes partly divided by a diffuse line of black scaling not extending to the basal
most part of the patch. An additional light patch occurs along the discal cell vein in cell
CuA2 which is roughly parallel to the basal discal cell patch, but extends slightly farther
distally and not as far basally. Ventral forewing markings similar to dorsal surface
except the light colored patches are larger, and narrow median band patches are also
present in cells R4 and R3. Also, there are distinct small off white patches distal to the
postmedian dots (not a clean pure white like the postmedian dots), usually in cells M3-R3
and progressively less often in cells CuA2 and CuAl. The light forewing markings have


10
evolutionary lineage. The point where an evolutionary lineage loses its ability to merge
with another lineage is in my view the theoretically and biologically significant event that
separates a species lineage from a lower lineage. Although Wiley (1992) argued that a
major advantage of his species concept was its applicability to all types of organisms,
with respect to Lepidoptera (sexually reproducing organisms) a comparison can be made
with the Biological Species Concept of Mayr (1963), where species were defined as
"Groups of actually or potentially interbreeding populations, which are reproductively
isolated from other such groups." For sexually reproducing organisms, the criteria of
"not actually or potentially interbreeding" and "maintains its identity from other such
lineages" are very similar, with the exception of "gray area" cases where some amount of
interbreeding occurs between two lineages yet they remain somewhat distinct. While
reproductively isolated lineages may remain very similar morphologically over long
periods of time, as exemplified by morphologically very similar butterflies and moths
isolated on different mountain ranges, for two separate lineages to reticulate into one,
reproduction must occur. In fact, with respect to sexually reproducing organisms, in my
view the evolutionary species concept is a consequence of putting the biological species
concept into an evolutionary context, achieved by adding a requirement that species units
be natural products of the evolutionary process. A very serious flaw of the biological
species concept is that it may recognize artificial (=nonmonophyletic) taxa not useful for
phylogenetic studies when applied to allopatric populations (Mishler and Donoghue
1992). Regarding allopatric populations, O'Hara (1994) points out a flaw common to
both the biological and evolutionary species concepts. In some situations, the decision of
whether two lineages are separate species is based on knowing what will happen in the


637
C. marina may be distinguished from C. erodyle by the dorsal forewing cream
colored median band of C. marina versus a sharp clean white median band in C. erodyle
erodyle (this character may be ambiguous for some C. marina females). In C. marina the
dorsal hindwing postmedian band is present in cells CuA2-CuAl and M2-M1, while in
C. erodyle it is usually confined to CuA2 or CuA2-CuAl and only rarely in M2-M1.
Also, the hindwing basal area is deep yellow and extending deep into the basal area in C.
erodyle erodyle, and cream or yellowish cream in C. marina which often has the deep
basal area black. More decisive characters occur on the legs and palpi. In C. marina the
ventral base of the labial palpi are white versus black in C. erodyle. Also, the tarsi of C.
marina contain only orange scales, while a mix of black and orange scales is present in C.
erodyle.
Males of C. marina may be distinguished from males of C. melitaeoides due to
their pale cream forewing median band, which is distinctly yellow in C. melitaeoides. In
females of C. marina the dorsal forewing median band is the same color as in males,
while in C. melitaeoides females it is a clean pure white. The shape of the forewing is
distinctly different between C. marina and C. melitaeoides, which is evident when series
are compared side by side. Males of C. melitaeoides have a more triangular forewing
with the apex narrower and less smoothly rounded off relative to C. marina. Females of
C. melitaeoides (only two were examined) have a much broader forewing than those of
C. marina, and the shape of the apex is not so narrow as males but distinctly less
smoothly rounded off relative to C. marina. The two C. melitaeoides females examined
have a complete red orange dorsal hindwing postmedian band patch and no postmedian
dot in cell M3, neither of which were ever found to be the case in C. marina. The


610
body color with no orange scaling". Subspecies leanira is polymorphic for the presence
or absence of orange scaling on the abdomen, but many specimens from the eastern part
of the range (closest to fulvia populations) have orange scales present on the abdomen.
Finally, Smith and Brock (1988) report an inverted "y" mark is absent in the discal cell in
subspecies leanira. This corresponds to character 96 in the phylogenetic analysis of the
preceding chapter, which serves as a synapomorphy for the C. leanira subspecies. In all
four subspecies, there may be a complete "y" or "m" connecting across the discal cell,
composed of a anterior component that extends basally into the discal cell and a shorter
posterior component. Also, in all four subspecies, the posterior component may be
absent or diffuse, and the anterior component may be diffuse. The latter is the case in the
specimen figured by Smith and Brock (1988), where I can make out the diffuse anterior
component in the discal cell of the right hindwing. By far the most common condition in
subspecies leanira is to have the anterior component sharp and conspicuous, but the
posterior component absent or diffuse. This condition is also common in subspecies
fulvia and cyneas, but these taxa have a much higher proportion of individuals with the
posterior component conspicuous as well.
Further Description (See also Further Description for species leanira (above) for
characters which are invariant among C. leanira subspecies): Orange and black scales
and hairs on the labial palpi with orange dominating. Dorsally palpi with predominately
orange scales but some black near the base and sometimes farther distally-most often in
specimens from the eastern part of the range; outer lateral side with the white lateral band
extending past the eye but not reaching the terminal segment, surrounded by orange
scales and hairs; ventrally covered with orange hairs and scales except white at the base


13
taxa based on the presence of evidence to suggest they are distinct lineages but the
absence of evidence to suggest they are evolutionary species. In other words, my
definition of a lepidopteran subspecies in practical application is: a monophyletic
lineage, allopatric with its closest relatives, with which it has identical genitalia (and
other sclerotized structures) but discontinuities in wing pattern. A subspecies could be
increased in rank to a species if evidence is acquired that favors the hypothesis tested to
delimit an evolutionary species over an alternative hypothesis.
Subspecies Concepts Rejected in this Work:
The vast majority of Chlosyne subspecies included by various authors do not meet
either my criteria for a valid taxon or a valid subspecies. A common method of
delimiting subspecies within the Chlosyniti has been to name populations, groups of
populations, or groups of individuals comprising parts of different populations,
representing points along a continuum of geographic variation. Geographic gradients are
often non-uniform, with phenotypes remaining more constant over some areas and
changing more rapidly over others. Subspecific status is assigned to populations
occupying various geographic areas based on the criterion (extremely arbitrary in my
view) that the worker(s) naming the subspecies feel it is sufficiently distinct to warrant
recognition (see quote by Austin and Smith 1998a above). Populations that represent
intermediates between the named populations are not assigned to any subspecies taxon,
and are referred to as subspecies A/B blends or intermediate populations (see Austin and
Smith 1998a, Austin and Smith 1998b) or as unassigned or undetermined populations
(see Smith and Brock 1988). Because some populations or individuals within the species
are omitted from the subspecies concepts, the subspecies concepts are by definition


531
Systematics of Western North American Butterflies: 141-142 & 155 (figures 9-
12). Type Locality: California: Tulare County, Kennedy Meadows Road, 16-20
MI W of Hwy 395. Holotype: Natural History Museum of Los Angeles County.
Note on Lectotype Designation: A lectotype has been designated from one of four
syntypes from Boisduval's collection by J. Emmel et. al (1998a) and is plated in their
figures 99-101. Miller and Brown (1981) restricted the type locality from California to
the vicinity of San Francisco; however J. Emmel et al. 1998d re-restricted the type
locality to Hwy 70 at Chambers Creek in Plumas County to correspond to a locality
where known populations exhibit geographic variation corresponding to the type.
Diagnosis: Chlosyne palla can be confused with C. hoffmanni, C. acastus, C. gabbii, and
C. whitneyi. As noted above, the male genitalia are distinct from C. hoffmanni plus there
are some helpful wing pattern characters. I find no consistent genitalic differences
between C. palla, C. acastus, C. gabbii, and C. whitneyi, and believe genitalic differences
reported within this group by Higgins (1960) were erroneous conclusions resulting from
inadequate sampling and/or style of preparation (compressing genitalia on slides).
Where C. palla and C. gabbii co-occur (western California), the ventral
background color is distinctly cream colored throughout in C. palla and distinctly white
distal to the postmedian band in C. gabbii. However, note that a portion of C. palla
specimens have the patches distal to the postmedian band a lighter or paler cream relative
to the remainder of the ventral background color. In C. whitneyi whitneyi the ventral
hindwing background color is pale white throughout with no cream scaling. C. whitneyi
damoetas may be separated from C. palla and all members of its clade by the orange and


776
PLATE J


529
extremely limited with the exception of Chlosyne whitneyi damoetas. Unambiguous
completely universally applicable identification characters are weak or lacking for all
taxa except C. whitneyi damoetas. Species delimitation is based partly on unpublished
hybrid inviability studies mentioned in Scott (1986) cited to J. Emmel. Reliable
determination of these taxa can be highly problematic, but specific data on location of
capture can be quite helpful. Hindwing background color distal to the postmedian band
was not coded as a phylogenetic character because within the Chlosyniti it is highly
variable not falling into clear discrete states. However, if this character were coded for
the C. hoffmanni clade and adjacent clades (C. harrisii clade and C. definita clade) the
cream color of C. hoffmanni and C. palla would be polarized as a symplesiomorphic
state, and the pearly white background color of C. gabhii, C. acastus, and C. whitneyi
would be polarized as derived, providing some weak evidence that these latter three taxa
may be more closely related to each other than to C. palla. However, such a character
would be ambiguous to score for some individuals even within the C. palla clade.
Another potentially better character is the length of the larval spines. Scott (1986) reports
that "the larval spines of C. acastus and C. palla are longer than those of C. gabbii and C.
whitneyi". If the larval spines of C. hoffmanni, the sister taxon to the C. palla clade, are
like either C. palla/C. acastus or C. gabbii/C. whitneyi, this character could be polarized
and provide a synapomorphy for either C. palla/C. acastus or C. gabbii/C. whitneyi.
Chlosyne palla (Boisduval)
Melitaea palla Boisduval, 1852. Ann. Soc. ent. Fr. (2) 10:275. Type Locality:


189


672
hindwing area is black dorsally (in contrast to forms lacinia and saundersi). There is an
orange patch on the dorsal hindwing occupying the distal ends of cells 1A+2A and CuA2
which is not distinct but continuous with orange in cell CuAl and subsequent anterior
cells. The ventral forewing has the same light markings as dorsally but most of them are
larger (some patches of the median band may be similar in size). Orange is present on
the distal side of the postmedian band patches (occasionally partially or even entirely
separated from the cream by black, but usually not) in cells CuA2-CuAl, and usually also
cells M3-M1 and sometimes R5. Light cream dominates the ventral hindwing with the
contrasting black symmetry system bands and black postmedian band. Orange patches
occur along the basal side of the black postmedian band in cells CuA2-Ml. The adjutrix
form is most similar to form saundersi, but it integrades to form crocale and has no
intermediates with saundersi. Gaps in the range of variation between adjutrix and
saundersi include the dorsal hindwing basal area black in adjutrix versus predominately
orange in saundersi, and the absence of differentiated orange and cream scales in the
dorsal hindwing median band in saundersi (all orange and sometimes with scattered
diffuse cream scaling). In some localities form adjutrix co-occurs with form crocale
and/or adjutrix/crocale intermediates, but apparently not other phenotypes. In
collections, form adjutrix is often confused with C. californica, especially individuals
with well developed dorsal hindwing patches distal to the postmedian dots.
Form crocale has a relatively narrow pale cream dorsal forewing median band,
with the patches appearing almost white. Light markings basal to the dorsal forewing
median band are usually absent or small or diffuse, with the exception of the discal cell
patch basal to the position of the discal spot which may be either prominent, small and


487
M. anomalus are identical to M. comeara. No other Microtia species have the dorsal
forewing surfaces exclusively with black and white scales, and M. comeara may readily
be separated from Chlosyne with only black and white on the dorsal wing surfaces
(Chlosyne lacinia "quehtala" phenotype, Chlosyne hippodrome, some forms of C.janais,
and C. rosita riobalensis) by the presence of diffuse orange bands on the ventral forewing
and hindwing between the positions of the marginal and submarginal bands, and by the
presence of orange between the bands of the central symmetry system. The male
genitalia of M. comeara and M. anomalus may be distinguished from other Microtia by
the somewhat hour glass-shaped plateau on the ventral juxta (Figure 113) and the very
shallowly notched saccus (Figures 112-113). The female genitalia (Figure 279) share
with M. dymas (Figure 281) the derived state of the sides of the ridge on the lamella
antevaginallis being greatly produced over the middle; however, they differ by having the
ridge less pronounced than M. dymas and by having a smaller and shorter ventral corpus
bursae plate.
Further Description: Labial palpi without orange hairs and scales. Dorsally labial palpi
with black scales and hairs only; outer lateral side surrounded by black hairs with white
ventrally-center and at base and black dorsally and at tip; ventrally with black scales and
hairs except white at base; inner lateral side with white scales and hairs dorsally and at
base and black scales and hairs ventrally and at tip. Vertex black with scattered white
scales. Tuft of white scales on the frontoclypeus anterior-lateral to the base of each
antenna well developed. Frontoclypeus lacking orange scales. Inner anterior-lateral
margin of the eye lacking an orange patch. Scales in the collar between the head and
pronotum black with white. Dorsally and laterally abdomen covered with black scales


788
PLATE U
17
§ 5 ^
JP'&Sf r
nr
pF*
k '
11
r i2
' *
r 13


44
three new states for character 18. An analysis with this alternative coding scheme did not
affect the resulting consensus tree generated by PAUP.
32. In lateral view, the orientation of the valvae:
0=Much more posterior than ventral (Fig. 76).
l=Much more ventral than posterior (All other figures of male genitalia capsules
in lateral aspect).
32.1. Outer lateral side of each valve with a groove, originating near the anterior-dorsal
comer of the valve and slanting posterior ventrally across the valve for much of its
length.
0=Absent.
l=Present (Figs. 60 & 64.1).
Characters of the juxta:
A number of Melitaeini have various ridges or plateaus on the ventral surface of
the juxta, while others have a predominantly smooth featureless juxta. The nymphalid
out groups have the juxta features quite different from the Melitaeini.
33. Ventral surface of the juxta:
l=With a mid ventral ridge extending to the posterior edge of the juxta and
originating a variable distance anteriorly (Fig. 5).
2=Fairly smooth and lacking distinct ridges or plateaus (Figs.
4,34,40,43,61,63,75,80,87,89 & 94).
3=Narrow ridge tapering anterior to posterior along the midline, flared anteriorly
into a broad triangular plateau (Fig. 37).


92
however, be equally parsimonious on the two aforementioned alternative topologies that
require two extra steps. Consequently, while the placement of Melitaeiti as the sister
clade to ((Chlosyniti,Poladryiti)(Gnathotrichiti,Phycioditi)) is two steps shorter than any
alternative topology, there are no convincing universal synapomorphies supporting this
topology.
A topology including ((Phycioditi,Gnathotrichiti)Poladryiti) requires three
additional steps over the most parsimonious tree with Chlosyniti and Poladryiti as sister
taxa. One character actually supports this topology, the sclerotization pattern on the
ventral surface of the phallus, which would require only four steps rather than five on a
tree where Poladryiti is the sister taxon to (Phycioditi,Gnathotrichiti). While three extra
steps is arguably notably less parsimonious, especially with respect to characters from
sclerotized structures, it must be noted that there is no universal synapomorphy from a
binary character supporting (Chlosyniti,Poladryiti) and only one universal synapomorphy
from a terminal derived state of a multistate character. This is the presence of a ventral
curved and pointed valve process with a convex anterior side and concave posterior side,
and the presence of a ventral valve process appears to have independently evolved in the
Melitaeiti clade. Consequently, while considerations of parsimony arguably provide
good evidence for (Chlosyniti,Poladryiti), considerations of character quality fail to
provide as convincing of evidence for monophyly, at least relative to the evidence of
individual monophyly for Chlosyniti and Poladryiti. This is why, based on Robbins and
Hensons (1986) third guideline, I conclude that by the criterion of stability, a subtribal
classification of Chlosyniti and Poladryiti is preferable to erecting one subtribe including
both clades.


755
USA:TX,NM,CA,OK
Mexico:Chihuahua,Sinaloa,Jalisco,Nuevo Leon,
San Luis Potosi (SLP),Tamaulipas,Veracruz,
Hidalgo,Oaxaca, Yucatan
Central Amer.: Extreme W. Panama (Darien).
South Amer: Trinidad,Venezuela,Colombia,
Peru,Ecuador,Paraguay,Bolivia,N. Argentina,
W. and extreme S. Brazil
SAUNDERSI
ADJUTRIX
South Am:
Colombia,
Venezuela,
Ecuador,
Bolivia.
Central America: N
El Salvador, Honduras,
Nicaragua, reappear
In W. Panama
SA: TX,NM,CA,NV.
UT. Mexico: SLP
lacinia/saundersi
integrades
adjutrix/crocale \
adjutrix/quehtala integrades \
integrades: \
Mexico: Veracruz, Tamaulipas, Yucatan,
San Luis Potosi, Campeche
USA: TX,NM,AZ,CA,NV,UT
Mexico: Sinaloa,Colima,Jalisca
Chihuahua
saundersi/quehtala \
integrades: \
W. Panama from \
Rio Armila to Darien
Honduras & Nicarag
CROCALE
LACINIA
Mexico: Veracruz, San Luis Potosi
Chiapas,Morelos,Oaxaca.
Central America: Guatemala,El Salvador,
Honduras,Costa Rica, Panama
iPAUPERA,
lacinia/quehtala
integrades
crocale/quehtala
integrades \
Mexico: Veracruz, Guerrero twS
Morelos, Oaxaca.
Central America: Guatemala,
El Salvador,Honduras,Costa Rica,
Panama
Mexico: Sinaloa,Colima,
Jalisco, Guerrero, Michoacan
Mexico: Sinaloa,Colima, Jalisco, Veracruz,Guerrero,Morelos
Michoacan. Central America: Reappears in Costa Rica,Panama
QUEHTALA
Figure 345: Diagrammatic representation of the continuum of geographic variation exhibited by Chlosyne lacinia. The capitalized names in boxes designate
Phenotypes/forms along a continuum of variation; they do not designate subspecies.


581
distal edge of the dorsal forewing has the postmedian band, postmedian dots, and patches
distal to the position of the discal parafocal elements compressed into a much smaller
area than for most taxa, including C. theona and C. chinatiensis. The postmedian band
patches are orange, and most developed in cell CuA2 and progressively less developed in
cells CuAl, M2, Ml and are reduced to sparse diffuse scaling or absent in other cells.
The cream colored postmedian dots appear to be absent in one specimen, occur in cells
M2-R3 (minute in M2 and Ml) in another, and occur in fewer cells in the remaining two
indicating the same range of variation characteristic of other taxa in the C. ezra clade.
The postmedian dot in cell R5 is distinctly larger than any other postmedian dots. The
patches distal to the position of the distal parafocal elements are about the same shade of
cream as the median band and postmedian dots, and the largest patch is present in cell
M3 where it extends basally into the position where a postmedian band patch would
occur (a postmedian patch is absent in this cell in all specimens examined); an additional
small sharp patch occurs in cell M2 in all specimens examined, and in one representative
small but conspicuous patches occur in M1-R5 while in other specimens they are diffuse
or absent in these cells (the range of variation falls within that exhibited by C. theona for
the homologous patches except the cell M3 patch does not extend basally to the position
of the postmedian band in C. theona).
The ventral forewing surface is similar to the dorsal forewing surface, and all of
the light markings present dorsally are present ventrally as well. However, the orange
postmedian band is much more prominent ventrally, with conspicuous orange patches in
cells CuA2-CuAl and M2-R3 plus sometimes with a small patch in cell M3 at the basal
edge of the expanded cream patch (noted above). The cream patches distal to the


259
vertically (white tufts ventro-lateral of antennae base are coded as a separate character
and not included here):
l=White patches of scales touching eyes ventro-laterally of antennae bases, white
patch centered on ventral side of the face (the extent of this patch is often highly
variable intraspecifically), black scales and hairs elsewhere.
2=Face with predominantly black scales and hairs with sparse white hairs usually
present (broken off when absent?) on the ventral part of the face.
3=White patches touching eyes ventro-lateral of antennae bases, white vertical
stripe centered on face below midpoint of area between antennae, black scales and
hairs elsewhere.
4=White vertical stripe centered on face below midpoint of area between
antennae, white patch centered on ventral side of face, black scales and hairs
elsewhere.
5=Exhibits a broad range of variation as follows (but the variation does not
overlap with other taxa): Small patches of white scales touching eyes around
white tufts vento-lateral of antennae (non variable), white stripe/band (thin to
broad) ventral to midpoint of area between antennae, white patch centered on
ventral side of face, ventral white patch may fuse with vertical stripe/band such
that the two are indistinguishable, entire face may be predominately white or with
black scales and hairs between white markings.
*6=AU black or black with an orange ventral patch, no white scales and hairs.
@*A=0&1: Some individuals are state 0 while others are state 1.


326
3=Present except the distal part of the fork in the discal cell is absent OR like this
plus the portion of the band posterior to the discal cell is absent (see below for an
explanation of this coding).
@A=0&1: Individuals may be either state 0 or state 1, or have sections of the
band present such that the completeness of the band is intermediate between state
0 and state 1 (contrast below).
@*B=0&2: Varies along a continuum between state 0 and state 2.
The two forms of state 3 were combined into one state for the following reason:
separating them into two distinct states would create two autapomorphic states providing
no phylogenetic information from this character for the two taxa involved. However, the
loss of the distal fork of the band in the discal cell but an otherwise complete band
anterior to this is a derived feature both of these taxa share.
Chlosyne leanira fulvia, C. leanira cyneas, and C. leanira cynisca lack any trace
of the above band in any individuals examined. However, this is the case in some but not
all specimens of C. leanira leanira, which most often has part of the band present,
especially in cell Sc+Rl and the discal cell. This is the only taxon for which coding this
character was problematic (other than those which have the part of the wing where the
band occurs all black). The variation in C. leanira leanira includes integrades from no
trace of the band to a very diffuse complete black band to a black band with a solid center
but diffuse edges in the appropriate position. Also, many individuals have sharp sections
in cell Sc+Rl and in the discal cell but with traces of the symmetry system absent in
other cells. I code this taxon as 0&2. All other taxa in the Thessalia group display a
range of variation including 0 or 1, with some intermediates (but never like state 3).


358
taxa C and D have acquired state 1, the most parsimonious arrangement for these three
taxa requires one step: that the ancestor which gave rise to C and D acquired state 1. The
theoretical preference of the topology (B(CD)) is that for a taxon to acquire a heritable
derived state, a particular mutation or series of mutations must occur, and that mutation
or series of mutations must spread in the population comprising that taxon. Since
mutations are thought to occur at random, most mutations decrease fitness, and even
mutations which improve fitness have a good chance of being lost due to chance before
becoming fixed in the population (Bell 1997) theoretically it is more probable that state
one evolved once in the ancestor which gave rise to C and D as opposed to having
independently evolved twice.
Now, consider the same scenario where the ancestor of three taxa had state 0 and
one extant taxon (B) has retained that state, except now taxa C and D include individuals
with either state 0 or state 1, or C & D have individuals which exhibit a range of variation
from state 0 to state 1. While state 1 has not become fixed in the population of C or D,
just as in the preceding example I would argue that a consistent application of parsimony
favors the topology (B(CD)). The most parsimonious arrangement is that the ancestor
which gave rise to B and C acquired the mutation which produced state 1, and that C and
D diverged from their ancestor before this mutation became fixed (or that the mutation
never became fixed due to changes in selective pressures), as opposed to C and D
independently acquiring the same mutation. This alternative, that state 1 independently
evolved twice, in my view requires an extra step, no less so than for the preceding
example. Yet, as indicated in scenario 1 of Figure 337, PAUP considers a topology
(B(CD)), requiring that state 1 evolved only once, to be equally parsimonious with


558
forewing background is predominately orange (subspecies pardelina-see below). In
subspecies C. endeis endeis, the ventral forewing is only orange in the discal cell and
immediately adjacent areas, unlike C. definite! where almost the entire ventral forewing is
orange. On the dorsal forewing surface, C. definita has six or seven patches composing
the hindwing postmedian band, whereas C. endeis usually has only four (CuA2-CuAl
and M2-M1), but some specimens of endeis pardelina have five (including cell M3).
Ventrally, the hindwing postmedian band occupies cells CuA2-Sc+Rl in C. definita and
CuA2-Ml in C. endeis (the patches are often not all the same color in either taxon, and
the patch in M3 is often reduced in both taxa). Dorsal forewing between the position of
the submarginal band and distal parafocal elements with sharp patches in at least some
cells in C. definita, while this area is predominately black with only very diffuse patches
in C. endeis. Also, the dorsal forewing postmedian dots are cream colored in C. definita
and white in C. endeis. Chlosyne marina may be separated from C. definita by many
characters, but some of the most obvious include the absence of orange scales on the
labial palpi and the absence of detectable discal spots and discal cell patches with orange
fill on the dorsal wing surfaces. Also, both C. endeis and C. marina tend to be larger
butterflies with less elongate wings relative to C. definita.
Further Description: Both orange and black scales and hairs on the labial palpi. Dorsally
palpi with predominately orange and scattered black scales and hairs; outer lateral side
with a white lateral band extending distal to the eye (but not reaching the terminal
segment) surrounded by orange hairs and scales; ventrally with predominately orange
scales on the inner side and at tip (the tip may also contain variable amounts of black)
and predominately dark hairs (with some white hairs) over white scales on the outer side


166
ny>.
j.A J \
\ \ / -V l\ \
V W:1
Figures 114-121. Male genitalia capsules of Microtia species. 114: M. eleda (lateral). 115: Same
(posterior inverted). 116: M. elva (lateral). 117: Same (posterior inverted). 118: M dymas (lateral).
119: Same (posterior inverted). 120: Microtia coracara (lateral). 121: Same (posterior inverted).


441
combination support the most parsimonious topology but by themselves are congruent
with many topologies. Many homoplastic characters are not mentioned but also
contribute to state changes at generic nodes. Also, some character states included under
terminal derived states of multistate characters are not unambiguous synapomorphies
despite occurring only in a terminal clade, due to ambiguity in the ancestral state
assignment at the node below the terminal clade and its sister clade. Character states
uniform within the Chlosyniti are included in the description of Chlosyniti but not
repeated in the generic descriptions unless some taxa could not be scored for the
character. A description of and key to all the subtribes of the Melitaeini, including the
Chlosyniti, was presented in Chapter 2 and is not repeated here.
Species/Subspecies Diagnosis: This section provides some of the characters most helpful
for separating a taxon from the taxa it most closely resembles. My work curating the
Chlosyne in the National Museum of Natural History and the Florida State Collection of
Arthropods provided insight into what taxa are most often confused, and I have made a
point to specify characteristics separating each taxon from taxa with which I have found
it confused in museum collections. In many cases, the diagnosis does not include all
characters which can separate similar taxa, but focuses on some which are easiest to
interpret and examine.
This section also includes genitalic characters which separate a taxon (or group of
taxa with identical genitalia) from taxa which are either similar and/or closely related.
With a few exceptions, reliable separation of Chlosyniti taxa does not require genitalic
dissection, and may be achieved with the pattern characters included first in the diagnosis
and in the key. Most of the genitalic characters included for the genus Chlosyne are those


156
Figures 61-62. Male genitalia capsule of Gnathotriche sodialis. 61: Ventral.
62: Dorsal.


428
Analysis
4 Genitalic Characters
l Pattern Characters
I All Characters
Figure 332: Variation in the proportion of homoplastic characters with different numbers of taxa in the
analysis, based on the Chlosyniti/Poladryiti data matrix (Table 5) analyzed with heuristic searches with
multistate taxa treated as polymorphisms in PAUP 4.0b4a.


202


265
2=White scales and hairs dorsally, thin band of orange scales on basal-lateral side,
and black hairs and scales ventrally.
3=White scales and hairs dorsally, black scales and hairs ventrally.
4=Covered with white hairs and scales throughout.
*5=Sexually dimorphic, females have state 0 and males have all black scales and
hairs.
*6=Covered with black scales only.
*7=Black scales dorsally, and a variable mix of white and black scales on the
outer lateral and the ventral sides.
Three taxa were found to be variable for this character, C. lacinia (this taxon exhibits the
greatest variation), C. hippodrome, and C. poecile. These taxa are coded ?.
74. Presence of black scales on the femur of the meso- and metathoracic legs.
0=Absent.
l=Present.
@*A=0&1: Geographically variable (see below).
One taxon is polymorphic for this character. The quehtala and lacinia phenotypes for C.
lacinia have black scales, but the adjutrix, crocale, and saundersoni phenotypes do not.
These taxa are coded dimorphically for this and the following character, and they appear
to exhibit a true dimorphism without intermediates. It is also possible that C. theona
perlula may be polymorphic for this feature; however, I had only four specimens, all of
which are partially denuded in this area, to examine when I scored this character. Two of
these have black scales on the femur whereas the other two appear not to. I code C.
theona perlula as "?". Also, note that populations of C. damaetas damaetas from


622
background, but some specimens which are very orange dorsally are ventrally identical to
specimens which are at the dark end of the continuum dorsally and vice versa. Cells
CuA2 and Ml to R3 distal to the postmedian band on the ventral forewing are
predominately cream with sparse if any orange while cells CuAl to M2 are
predominately orange in this area with sparse if any cream scaling. The ventral hindwing
pattern is similar throughout the range and varies from having a pale cream/white
background color to having a slight yellow tint within populations from both northern
and southern localities.
The variation in the amount of orange present on parts of the body appears to
partially depend on the amount of orange present on the dorsal surface of the wings in
both males and females. Specimens from the southern part of the range tend to have
fewer orange scales than northern specimens, but the full range of variation is exhibited
by both Kansas and Arizona specimens.
I have examined paratypes of the taxa designated as Chlosyne fiulvia pariaensis
(from Kane County, Utah) and Chlosyne fulvia coronado (from Pima County, Arizona)
by Smith and Brock (1988), and they are normal specimens within the continuum of
variation. One male paratype of coronado is the largest male I have seen, but the other is
well within the normal range of size variation for Apache County, Arizona specimens.
The name pariaensis refers to an orange phenotype resembling the alma phenotype of
subspecies leanira in wing pattern. The male paratype (from Kane Co. Utah) is the
orange phenotype that appears to occur throughout the range of C. leanira fulvia, and the
female is the extreme orange phenotype present in Kansas and Utah, but apparently not in
Arizona specimens from the areas I have examined (Smith and Brock 1998 report


124
narrow bridge (Fig. 2), vesica without paired lateral sclerotized plates attached to each
ventral lateral side of the phallus.
Females: Corpus bursae encircled by a continuous band (of nearly uniform width) of
inverted teeth (Fig. 261), ventral plate on corpus bursae distinctly separated by
membranous tissue along the ventral midline (Fig. 261), a distinct sclerotized plate
(originating at the posterior end of the ductus bursae at the junction with the ventral
genital opening) on the ventral ductus bursae separate from and posterior to the preceding
plate (Fig. 261).
Distribution: Holarctic Euphydryiti
1. Males: Inner projection of valvae lacking a posterior fork, inner wall of valvae with
only hair-like setae, medial area of saccus prominent (may be notched), vesica with
paired lateral sclerotized plates attached to each ventral lateral side of the phallus (Figs.
211-212).
Females: Corpus bursae not completely encircled by a band of sclerotized teeth OR the
band is not of nearly uniform width, ventral plate on corpus bursae not separated by
membranous tissue along the ventral midline, no additional sclerotized plate on the
ventral ductus bursae posterior to the preceding plate.
Distribution: Holarctic or Neotropical 2
2. Males: Ventral membranous part of everted vesica covered with minute teeth (Figs.
52 & 54-56) (Didymaeformia trivia is the only possible exception of the taxa examined),
teeth bearing sclerotized plates on the vesica (if present) elongate longitudinally with
many minute teeth spread over the surface and edges (Figs. 52 & 56-57).


602
Wing fringes with the usual black and white checkered pattern with the white
sections between the veins. Forewing basal costal area with a thin row of orange scales.
Many wing pattern characters are variable geographically within and/or between
subspecies.
Dorsal forewing median and postmedian bands cream colored, with a patch that is
at least partly cream at the distal end of the discal cell and in cell CuA2 in the vicinity of
the origin of vein CuA2. Discal spot absent. No postmedian dots within the postmedian
band. Other markings vary between and/or within subspecies. Some specimens have the
postmedian band composed of small dots of the same size as the clade of derived
Chlosyne with white postmedian dots, including all specimens of subspecies cyniscci.
However, subspecies leanira and cyneas show a continuum of variation between large
patches forming the postmedian band to the patches reduced to dots. The phylogenetic
analysis demonstrating the relationship between the four leanira species favors the
hypothesis that the dots in the postmedian area of subspecies cynisca are homologous to
the postmedian band wing pattern elements and not postmedian dots.
Dorsal hindwing with a cream colored median and postmedian band, although
patches of the postmedian may be so small as to appear like postmedian dots, but no
postmedian dots appear in the postmedian band patches and the postmedian band patches
vary along a continuum between prominent patches and small dots within all subspecies
except cynisca, where the patches are always small dots.
Ventral forewing with orange in costal cell continuous with surrounding orange
and not forming a distinct patch. Light markings present dorsally also present ventrally
where they may appear sharp or very washed out, excluding the postmedian band which


ACKNOWLEDGMENTS
I wish to acknowledge a number of people who provided valuable assistance with
this project. For serving on my doctoral committee, providing advice and assistance
throughout this project, and for review of this manuscript, I thank Thomas Emmel, James
Lloyd, Jonathen Reiskind, Frank Slansky, and John Heppner. For providing access to or
loan of specimens critical to this study I thank John Heppner (Florida State Collection of
Arthropods), Robert Robbins and Donald Harvey (National Museum of Natural History),
Lee and Jackie Miller (Allyn Museum), Keith Willmott, and Jason Hall. I am grateful to
Gerardo Lamas for sending me scanned images of several Chlosyne types in the British
Museum. For providing work space in the Florida State Collection of Arthropods 1 thank
John Heppner. Thomas Walker provided access to an exceptional camera lucida
microscope set up for doing the genitalia illustrations. I thank Steve Lasley and Nick
Hostettler for computer assistance. I also thank Thomas Emmel and the University of
Florida's Department of Entomology and Nematology (especially the graduate
coordinator, Grover Smart) for providing research and teaching assistantship funding,
respectively, for some semesters. Donald Harvey and Robert Robbins provided
considerable advice and assistance during a visit to the National Museum of Natural
History, and Robert Robbins assisted with securing a short term visitors grant for me to
visit that institution. Debbie Hall was most helpful for assistance dealing with the
University of Florida's bureaucracy. For helpful discussions and/or debates on species
problems, cladistics, and/or other issues in taxonomy I thank James Lloyd, Donald


72
The signa teeth are easier to detect, and detectable at a lower magnification, when the
genitalia are placed in a glass petri dish with water and a very bright light source from
underneath. There were some cases where scoring this character within the Phycioditi
was somewhat ambiguous, and examination of additional Phycioditi may warrant the
conclusion that this character should not be used in a phylogenetic study of that group.
89. For those taxa with character 88 state 2, the arrangement of the aggregations of small
teeth on the inner surface of the corpus bursae:
l=Continous band (of similar width) of teeth encircles corpus bursae (Fig. 261).
2=Distinct signa patches of teeth present on lateral sides, but connected by a
narrower bridge of teeth on the ventral side (in Melitaea cinxia and Mellicta
britomartis the bridge of teeth is broken at the ventral midline) (Figs. 262-265).
3=Teeth confined to two distinct patches on the lateral sides of the corpus bursae
(Figs. 269-270,278,280,282,284,286,287,290 & 291).
4=A continuous band of teeth which is widest on the lateral sides, narrow across
the dorsal side, and relatively wider (than dorsally) across the ventral side (Fig.
267).
5=Nearly continuous band of irregular width encircles the corpus bursae, except
for a narrow strip along the ventral midline which is nearly devoid of teeth (Figs.
274-276).
#6=The teeth densely cover the entire corpus bursae anterior of the ductus bursae.
90. For those taxa with character 88 state 2, the relative sclerotization of the teeth on the
corpus bursae:


436
CALEOS Model
A
2
E B
0&1 2
_.+l
_ -2X>
C
1
D
1
E
0&1
D
I
C
1
A
2
B
2
D
1
C
1
EA BAB EC D
0&1 2 2 2 2 0&1 1 1
. J2
4 Steps
..2>l
3 Steps
B
_.0>2
__+l
3 Steps
._+!
3 Steps
D
Distinct State Model
A
2
E
a
C
1
D
1
3 Steps
E D
a 1
l>a
C
1
A
2
B
2
D
1
A
2
B A
2 2
B
2
C
1
3 Steps
_ _2>a
3 Steps
__0>2
DPCWH Model
. _2>a
3 Steps
j0>2
H
_.0>2
E
0
1
_.a2X>
_ b:2>l
B C
2 1
2 1
D E
1 0
1 1
3.0 Steps
' a: 1>2
_Lb:l>2
D
1
1
a 1X>
c
1
1
A
2
2
B D
2 1
2 1
A
2
2
B
2
2
A
2
2
B
2
2
E C D
0 1 1
I 1 1
2.5 Steps
" a:2>l
_ b 2> 1
2.0 Steps
_a:0>2
b l>2
' a 0>2
2.0 Steps
.a 0>I
. a0>2
_b0>l
I 1
.b0>l J
. b:0>2 K
. b:0>2 ^
0
0
0
0
Scenario 4
M
PAUP &
Discrete
DPCWH
CALOS
State
1 2
1 2
1 2a 2b
OG1
0 0
0 0
0 0 0
OG2
0 0
0 0
o
o
o

A
1 2
1 2
1 2 2
B
1 2
1 2
1 2 2
C
1 1
1 1
1 1 1
D
1 1
1 1
1 1 1
E
1 0.1
1 a
1 0 1
Figure 340: Behavior of different models for polymorphic characters on four tree topologies for
Scenario 4, with the polymorphic character mapped on the trees. A-D: CALEOS Model. E-H: Distinct
State Model. I-L: DPCWH Model. M: Data matrix for scenario 4. The polymorphic character requires
four steps on tree A and three steps on trees B-D in PAUP 4.0b4a.


5
taxon, which represents a product of the human imagination with no real existence in
nature. In other words, natural taxa are monophyletic evolutionary lineages, including by
definition all the descendents of a common ancestor. Artificial taxa are any groupings of
organisms which a human being artificially creates, and do not represent an evolutionary
lineage. I make no distinction between paraphyletic and polyphyletic taxa. A
paraphyletic group has been defined (originally by W. Hennig) as a group not including
all of the descendents of a common ancestor (Scotland 1992a), while a polyphyletic
group has been defined as "a group in which the most recent common ancestor is
assigned to some other group and not to the group itself' (Scotland 1992a, quoted from E.
Wiley via J. Farris). Because all life is believed (based on current evidence) to have a
common origin, any conceivable grouping of extant organisms that is not monophyletic is
by definition paraphyletic. Consequently, for the purposes of classifying extant
organisms in a natural classification scheme, the only true distinction is between
monophyletic (natural) and nonmonophyletic (artificial) groups.
Artificial classification schemes can be very useful for certain purposes, such as
deciding what to include in a book, what to include in a university course, what to include
in a pest control manual, etc. Some examples of paraphyletic groups sharing some type
of similarity used for these purposes include reptiles, invertebrates, moths (excluding
butterflies), the Lepidoptera of Florida, Florida pests of citrus, medically important
insects, urban pests, beautiful moths, etc. Although I recognize the utility of such
groupings for these purposes, I argue that such groupings (or any nonmonophyletic
grouping) have no place being formally named or ranked in the scientific discipline of
biological systematics for several reasons: 1) They are artificial, do not exist in nature,


160
48(2).
3(1),4(2)
55(17
Figures 74-79. Male genitalia of Phyciocles tharos. 74: Capsule (ventral). 75: Same
(dorsal). 76: Same (lateral). 77: Phallus (dorsal). 78: Phallus (lateral, with vesica
partially everted). 79: Phallus (ventral).
48(2)
55X


172


237
OConvex (Figures 108,110,112,136,139-145,147-150 & 154-155).
1 Concave (Figures 81,94,99,106,134-135,137-138, & 151-153).
2=The ventral edge is distinctly concave (Figure 148-this state is difficult to
distinguish from state 1 with a two dimensional Figure).
3=Neither distinctly convex nor concave (Figures 87 & 89).
@A=0&2, including state 0 and state 2 with intermediates (Figure 146).
The exact angle from which the male genitalia are viewed affects the apparent shape of
the inner valve process. In ventral aspect, as the anterior end of the genitalia are pushed
down (increasing the posterior aspect of the view) the posterior edge begins to look
concave including for state 0 taxa. The inner prong appears convex dorsally in posterior
view for all taxa. The state differences appear to be due to differences in how the base of
the inner valve process has been rotated, with state 2 taxa between the orientation of the
process of state 0 and state 1 taxa. The dorsal edge of state 2 taxa appears homologous to
the anterior edge of state 0 taxa and the posterior edge of state 1 taxa. Evidence for this
hypothesis is provided by one taxon, C. leanira, which exhibits state 0, state 2, and
intermediates. The range of variation is not geographic and occurs in all subspecies of C.
leanira, including C. /. leanira, C. l.fulvia, C. 1. cyneas, and C. /. cynisca. Consequently,
all of these taxa are coded "0&2". No other taxa were found to exhibit intermediates
between any of the above states, or to posses more than one state.
24. Extent of curvature of the entire inner valve process:
0=Strongly curved.
l=Fairly straight, only slightly curved (Figures 87,89 & 152-153).


623
pariaensis from extreme northwest Arizona, an area for which I have not seen
specimens). The two male paratypes of coronado are the extreme cream form and an
intermediate cream and orange form that occur throughout the range, although the latter
specimen is unusually large for a male in nature (the specimen was reared). The paratype
female is on the dark end of the continuum of variation for females.
Distribution: I have examined specimens of C. leanira fulvia from Arizona, New
Mexico, Colorado (few, none from the eastern part of the state), Utah (only Kane County
near the Arizona border), southeast Nevada (Lincoln County only) northwest Oklahoma
(Woodward County), Kansas, and west Texas. Data for specimens in the FSCA and
NMNH are presented in Table 12. A very detailed distributional map is provided by
Smith and Brock (1988), showing both records of subspecies fulvia and subspecies
leanira. It is interesting that Stanford and Opler (1993) indicate records for Kane
County, Utah and Lincoln County, Nevada for both subspecies leanira and fulvia (ranked
separate species), and also indicate records for both taxa from the northwestern most
Arizona county (as do Smith and Brock (1988)). The specimens offulvia I examined
from Kane and Lincoln counties were the typical fulvia phenotype of the northwest part
of the fulvia range (although one reared male from Kane County was larger than most
specimens offulvia that I have seen), with no evidence of greater tendency toward the
leanira phenotype. Smith and Brock (1988) noted that while subspecies fulvia and
leanira occur in the northwestern most Arizona county, they do not fly together in the
same localities (the subspecies leanira records are farther to the west). There are two
specimens in the National Museum of Natural History labeled "La Port Cal. Aug. 8-15
Barnes", well inside the range of C. leanira leanira, which I view with great suspicion,


411
Strict
C. nycteis
C. gorgone
C. n. sp. nr. har.
C. harrissii
C. hoffmanni
C. palla
C. gabbii
C. acastus
C. w. damoetas
C. w. whitneyl
C. definita
T. ezra
T. theona
T. perlula
T. chinatiensis
T. I. leanira
T. I. fulvia
T.cyneas
T. cynisca
C. e. pardelina
C. e. endeis
C. melitaeoides
C. marina
C. e. poecile
C. e. erodyle
C. melanarge
C. eumeda
C. cal ¡fornica
C. lacinia
C. ehrenbergi
C. hippodrome
C. narva
C. g. gaudealis
C. g. wellingi
C. janais
C. marianna
C. rosita rosita
C. montana
C. rosita browni
C. riobalensis
C. mazarum
T. eleda
T. coracara
T. anomalus
D. dymas
M. elva
I A. pelops
' A. proclea
P. minuta
H. fasciatus
H. miriam
A. perezi
A. pantoni
A. tulita
CumPhyGnaMel
Figure 316: Strict consensus tree from a heuristic search of successively weighted characters from the
Chlosyniti/Poladryiti data matrix, with multistate taxa assigned discrete states.


785
PLATE R
26
27
28
29
30


257
l=Absent.
The tuft tends to be very reduced in C. melnarge, and may appear absent in partly
denuded specimens.
All taxa have black hairs and scales on the frontoclypeus dorsal to the origin of the labial
palpi, and all have some white hairs with the exception of Microtia elva. The presence of
orange scales in this area varies within the Chlosyniti, but not within taxa, except when
part of a sexual dimorphism. Females tend to have more extensive orange scaling than
males for those taxa with orange scales in this area.
64. Orange scales and hairs on the frontoclypeus in males.
0=Present.
l=Absent.
@*A=0&1: Some individuals are state 0 while others are state 1.
65. Orange scales and hairs on the frontoclypeus in females.
CNPresent.
l=Absent.
@*A=0&1: Some individuals are state 0 while others are state 1.
The pattern of scales on the frontoclypeus is intraspecifically variable within some taxa,
but shows little variation and can be placed into discrete states for other groups of taxa.
Those taxa that posses both orange and white scales in this area tend to have variable
patterns which cannot be assigned into discrete states. These taxa are coded as
Those taxa lacking orange scales in this area have more consistent patterns, with the
exception of some forms of Chlosyne lacinia and Texola coracara. The variation of C.
lacinia is unlike the states for other taxa, so this problem is resolved by assigning C.


537
satisfactory", but decided to refer whitneyi to a form of palla based on "the presently
generally accepted interpretation". Higgins noted from Behr's (1863) original description
that the original specimens of whitneyi came from what is now Yosemite National Park at
the headquarters of the Tuolumne River in "elevated and uninhabited regions". Higgins
(1960) noted having seen no C. palla from that region, and questioned how well C. palla
fit Behr's (1863) description of C. whitneyi. Scott (1986) applied the name whitneyi to
malcom Comstock and damoetas (Skinner) (as an allopatric subspecies of whitneyi) but
gave no explanation for doing so, and his figure of nominate whitneyi appears to me to
actually be C. palla and not the malcom phenotype. J. Emmel et al. (1998c) also apply
the name whitneyi to what had been known as malcom, report that C. palla is not known
from the elevated areas near the headwaters of the Tuolumne River, and also include
Behr's (1863) original description, including the English part and Latin translation.
I somewhat concur with the view of these author's that Behr's (1863) description
is a better match with malcom than with palla. In particular, I note Behr's (1863)
statement regarding whitneyi compared with Behr's concept of palla "The striking
alteration in the colors of the upperside in the series of M. palla does not exist here".
Series of malcom which I have examined have very little contrast between the
postmedian band and the median band and other light forewing markings, while the
contrast is often (but not always) marked in C. palla. Another character mentioned by
Behr is "the crescents of the submarginal band [I interpret this as homologous to what I
term the postmedian band, as this is the band which contains eyespots/postmedian dots]
entirely lacking eyespots." While this character state does not apply to either palla or
malcom or any similar taxa, the postmedian dots are notably paler and less contrasting in


308
0=Absent.
l=Present.
The color of the ring of light colored scales ranges from cream to light orange in
Chlosyne nycteis but is always cream in Chlosyne gorgone. Since these are the only two
taxa with this feature, an additional character for color would result in two autapomorphic
derived states, or merely weight this character twice (depending on which model for
multistate taxa was used).
Taxa which lack hindwing postmedian dots are coded 0, because the lack of
postmedian dots does not preclude scoring if the ring of light colored scales are present in
the equivalent position (it turns out this is never the case in taxa lacking hindwing
postmedian dots).
Many taxa in the Chlosyne group of genera have two bands of wing pattern
elements between Nijhouts (1991) unnamed element g and the distal band of the central
symmetry system on the ventral hindwing surface. These bands, when complete, are
formed from a series of lines or patches in cells 1A+2A through Sc+Rl. However, in
some specimens the lines do not occur in all of these cells, and this variation can occur
intraspecifically. Nijhouts (1991) Nymphalid Ground Plan does not include these wing
pattern elements, and I name them as the distal median line and the basal median
line."
115. The presence of distal and basal median lines on the ventral hindwing as described
above:
0=Both a distal and basal median line are present.
l=Median lines are absent.


650
mostly continuous and jagged/uneven. Dorsal and ventral hindwing and forewing margin
between the position of the marginal and submarginal bands solid black. Dorsal forewing
patches comprising the median band yellow and concolorous with the dorsal hindwing
median band/basal area in subspecies poecile but contrastingly white in subspecies
erodyle. The dorsal forewing basal area contains yellow except for cells 1A+2A and
sometimes CuA2, and the yellow extends deep into the basal area. There is no distinct
hindwing median band separated from the basal yellow. Number and extent of black
symmetry system markings within the basal yellow on the dorsal hindwing variable, but
always less than ventrally. Dorsal hindwing between the position of the submarginal
band and distal parafocal element with very small and diffuse areas of cream scaling
usually not apparent without magnification. More prominent cream patches present in
the same position ventrally, but in many specimens the patches are small and only sharp
in cell M3 or M3 and adjacent cells, and in some specimens all patches are diffuse.
Patches formed between the distal and basal elements of both the basal and central
symmetry systems on the ventral hindwing coalesced together but with some sections
incomplete and disjunct. The bands of the distal and basal symmetry system are
distinctly separate (never fused together). The bands of each symmetry system have
black between them, but for the central symmetry system the black broken by a variable
amount of cream background color in most and often all cells. The size of sections of the
central symmetry system bands and the extent to which they are fused together is highly
variable in both subspecies. Ventral hindwing background color between the proximal
band of the central symmetry system and distal band of the basal symmetry system


452
22 Dorsal surface of labial palpi with predominately orange scales except at base
Chlosyne leanira leanira
22. Dorsal surface of labial palpi with black scales only 23
23. Dorsal wing surfaces almost always with at least some orange scaling distal to the
postmedian bands, ventral forewing predominately orange, or if not the orange is not
limited primarily to the discal cell and along the distal edge of the wing) 24
23. Dorsal wing surfaces completely devoid of orange scaling, ventral forewing with
orange limited to the discal cell (may slightly extend to adjacent areas) and along the
distal edge of the wing Chlosyne leanira cynisca
24. Dorsal hindwing basal to postmedian cream patches often predominately black with
sparse if any orange scaling between the median and postmedian bands, if not then none
of the patches distal to the postmedian band have cream scales, dorsal wing surfaces
never predominately orange throughout, abdomen lacking orange scales
Chlosyne leanira cyneas
24. Dorsal hindwing basal to postmedian cream patches varies continuously from
predominately orange to black but always with some conspicuous orange and/or cream
scaling between the median and postmedian bands, if specimens are not predominately
orange dorsally, cream scaling is present in some forewing patches distal to the
postmedian band, abdomen with at least some orange scales in fresh specimens
Chlosyne leanira fulvia
25. Area between the distal and proximal bands of the central symmetry system orange,
black scaling along veins broken at the basal end of veins R5 and Ml, frontoclypeus and
abdomen with orange scales 26


569
cell CuA2. The basal and distal bands of the symmetry system are black continuous with
solid black between them, except in the discal cell fork where either a cream or orange
patch or continuous solid black may occur between the fork (the fork is not distinct if this
area is solid black). The width of the symmetry system band is variable, especially in the
discal cell where the sides of the fork are quite narrow in some specimens and broad in
others. A prominent orange patch occurs basal to the H vein in the costal cell. Ventral
hindwing postmedian band patches are more prominent than in the same individual
dorsally, and the complete set of eight patches is present although the 1A+2A patch is
minute and diffuse in some individuals. The cream patches distal to the postmedian band
are conspicuous (with sharp or diffuse borders) in contrast to the dorsal surface, and
occur in cells CuAl to Sc+Rl and sometimes CuA2.
Geographic Variation: I find no evidence of geographic variation, and only a small range
of intraspecific variation as noted for some characters in the above description.
Range: I have seen specimens from Costa Rica, Panama, Peru, and Venezuela (specific
locality data is lacking for Peru and Venezuela specimens). Table 12 includes data from
the NMNH and FSCA. Higgins (1960) also reported records for Colombia, including
Cabaco Is. and Cauca. The type of Eresici yorita (reported to be a typical specimen of
ezra by Austin and Smith (1998a)), is supposedly from Honduras, but in my view
corroborating evidence should be acquired before this suspicious locality record is
accepted as valid.
Species Delimitation: Higgins (1960) considered ezra to be a subspecies of Chlosvne
theona, and chinatiensis (reportedly not actually seen by Higgins in Higgins (I960)) and
perlula to be forms o theona. Higgins (1960) stated with respect to both leanira and


Table 1-Continued
133
Characters
Taxa
41
42 43 44 45
46 47 48
49
50
51
CM
in
53 54 55
56 57
58 59 60
Chlosynel
1
0
1
1
1
1
1
1
0
0
0
0
1
2
1
1
1
?
7
2
Chlosyne2
1
0
1
1
1
1
1
1
0
0
0
0
1
2
1
1
1
7
7
2
*Texola elada
1
0
1
1
1
1
1
1
0
0
0
0
1
2
1
1
4
7
0
2
Microtia elva
1
0
1
1
1
1
1
1
0
0
0
0
1
2
1
1
4
?
1
2
Dymasia dymas
1
0
1
2
1
1
1
1
0
0
0
0
1
2
1
1
4
7
1
2
Texola corocara
1
0
1
2
1
1
1
1
0
0
0
0
1
2
1
1
4
?
1
2
Texola anomalus
1
0
1
1
1
1
1
1
0
0
0
0
1
2
1
1
4
?
1
2
Antillea pelops
1
0
1
1
1
2
1
1
0
0
0
0
1
1
1
1
2
2
0
2
Antillea proclea
1
0
1
1
1
2
1
1
0
0
0
0
1
1
1
1
2
2
0
2
Polydryas minuta
1
0
1
1
0
1
1
1
2
0
0
2
0
0
1
1
6
?
0
2
Higginsius fasciatus
1
0
1
1
0
1
1
1
2
0
0
3
0
0
1
1
3
7
0
2
Higginsius miriam
1
0
1
1
0
1
1
1
2
0
0
3
0
0
1
1
3
?
0
2
Atlantea tulita
1
0
1
1
?
1
1
1
2
0
0
4
0
0
1
1
3
?
0
2
Atlantea pantoni
1
0
1
1
7
1
1
1
2
0
0
3
0
0
1
1
3
7
0
2
Mellicta athalia
1
0
1
1
0
1
1
1
2
0
0
1
0
0
1
1
2
1
0
2
Mellicta britomartis
1
0
1
1
0
1
1
1
2
0
0
1
0
0
1
1
2
1
0
2
Mellicta aurelia
1
0
1
1
1
1
1
1
0
0
0
0
0
0
1
1
2
1
0
2
Mellicta varia
1
0
1
1
1
1
1
1
0
0
0
0
0
0
1
1
2
1
0
2
Mellicta parthenoides
1
0
1
1
0
1
1
1
2
0
0
1
0
0
1
1
2
1
0
2
Mellicta asteria
1
0
1
1
1
1
1
1
0
0
0
0
0
0
1
1
2
1
0
2
Mellicta alatauica
1
0
1
1
0
1
1
1
2
0
0
0
0
1
1
2
1
0
2
Cinclidia phoebe
1
0
1
1
1
1
1
1
0
0
0
0
0
0
1
1
2
1
0
2
Cinclidia scotosia
1
0
1
1
1
1
1
1
0
0
0
t 0
0
nr
1
1
2
1
0
2
Melitaea cinxia
1
0
1
1
1
1
1
1
0
0
0
nr
nr
0
1
1
2
1
0
2
Melitaea diamina
1
0
1
1
1
1
1
1
0
0
0
0
0
0
1
1
2
1
0
2
Didymaeformia didyma
1
0
1
1
0
1
1
1
4
0
0
0
0
0
1
1
2
1
2
2
Didymaeformia trivia
1
0
1
1
0
1
1
1
0
0
0
0
0
0
1
1
2
1
2
2
Phyciodes tharos
1
0
1
2
0
1
1
2
0
0
0
0
0
0
1
1
3
?
0
2
Phyciodes batesi
1
0
1
2
0
1
1
2
0
0
0
0
0
0
1
1
3
7
0
2
Phyciodes phaon
1
0
1
2
0
1
1
2
0
0
0
0
nr
0
1
1
3
?
0
2
Phystis simois
1
0
1
1
0
1
1
1
0
0
0
0
0
0
1
1
3
?
0
2
Anthanassa texana
1
0
1
2
0
1
1
1
0
0
0
0
0
0
1
1
3
?
0
2
Anthanassa tulcis
1
0
1
1
0
1
2
1
0
0
0
0
0
0
1
1
3
?
0
2
Telenassa teletusa
1
0
1
2
0
1
1
1
0
fcT
0
0
0
0
1
1
3
?
0
2
Castllia castilla
1
0
1
1
0
1
1
1
0
IT
0
0
0
0
1
1
3
?
0
2
Dagon catula
1
0
1
2
0
1
1
1
0
0
0
0
0
0
1
1
3
?
0
2
Ortilia llriope
7
0
1
1
0
1
1
1
0
0
0
0
0
0
1
1
3
7
0
2
Ortilia ithra
7
0
1
1
0
1
2
1
0
0
0
0
0
0
1
1
3
?
0
2
Eresia eunice
1
0
1
1
0
1
2
1
0
0
0
0
0
0
1
1
3
?
0
2
Eresia frisia
1
0
1
1
0
1
2
1
0
0
0
0
0
0
1
1
3
?
0
2
Janatella leucodesma
7
0
1
1
0
1
2
1
0
0
0
0
0
0
1
1
3
7
0
2
Janatella fellula
7
0
1
1
0
1
?
3
0
0
0
0
0
0
1
1
3
7
0
2
Mazia amaznica
1
0
1
1
0
1
2
1
0
0
0
V
0
0
1
1
3
?
0
2
Tegosa claudina
1
0
1
2
0
1
2
1
01
0
0
0
0
0
1
1
3
7
0
2
Tisona saladellensis
1
0
1
1
0
1
2
1
0
0
0
0
0
0
1
1
3
7
0
2
Gnathotriche exclamationis
2
2
1
1
0
1
1
1
0
0
0
0
0
0
1
1
3
?
0
2
Gnathotriche sodialis
2
2
1
1
0
1
1
1
0
0
0
0
0
0
1
1
3
7
0
2
Gnathotrusia mundina
2
1
1
1
?
_L
1
1
3
0
0
0
0
0
1
1
3
7
0
2


379
informative; 2) The most parsimonious distributions of some character states support the
best evolutionary hypothesis while others do not, hence not all character states are
equally informative; and 3) Almost half of the nonhomoplastic character states are states
of homoplastic characters, hence the relative value of the information content of different
character states cannot be determined from measures of overall character consistency. I
argue that in this case study these three observations provide experimental evidence
favoring a model for successive weighting of character states over successively weighting
entire characters. Furthermore, if either of the assumptions of a model for equally
weighting characters or successively weighting characters were true, a successive state
weighting model would perform just as well because the assumptions of that model
would also be true.
The original model for successive weighting of characters was based on character
consistency indices (Farris 1969), and later refined by using the rescaled consistency
index (CI*RI) in (Farris 1988). Since the retention index is in essence a measure of the
proportion of similarities on a phylogenetic tree determined as synapomorphy (Farris
1989), an equivalent of the rescaled consistency index cannot be used for weighting
character states since some character states represent symplesiomorphies. Rather, the
successive state weighting model is based on the assumption that states requiring extra
steps are less stable over evolutionary time and are consequently proportionately less
reliable indicators of evolutionary history.
I do wish to point out that I see two potential sources of error with the successive
state weighting model. One is that the character state which accounts for a homoplastic
step can be ambiguous. For example, consider the tree topology (1(2(3,4))) where 1-4


287
I am skeptical that this character and the presence of light scaling inside the discal
spot are completely independent, although some taxa are coded differently for these two
characters. There are numerous cases, however, where the discal spot or the basal
symmetry system patch (but not both) are detectable in a taxon; hence, it seems best to
include the presence of light scaling inside these two features as two separate characters.
95. Ventral forewing discal cell with a prominent patch of dark red-orange, sometimes
extending into surrounding cells:
0=Absent.
1 =Present.
96. In the ventral hindwing discal cell, the presence of a black line extending from near
the basal origin of the wing to the end of the discal cell, where this line forks and curves
distally and then back basally, forming a m-shaped to y-shaped bridge between the base
of the M3/M2 vein fork and the base of the M1/R5 vein fork:
0=Absent.
l=Present (but not always complete).
The sharpness or diffuseness as well as the completeness of this pattern varies within
taxa. In some individuals the base of the pattern is absent and only the m-shaped bridge
is apparent, in others the m-shaped bridge is apparent with the base absent, or the anterior
portion (including the base) is present but the posterior portion is absent. Also, note that
in some specimens the fork in this pattern comes well below the end of the discal cell,
practically at the basal base of the black line.


267
*4=Covered with black scales except for an orange area at the proximal end of the
tibia.
77. Scales on the tarsi of the meso- and metathoracic legs. Note that the ventral surface
of the tarsi is usually largely devoid of scales.
0=With a mix of orange and white scales, orange scales concentrated dorsally,
white scales concentrated ventrally.
l=With orange scales only.
2=With black scales only.
3=A variable mix of black and orange scales, often with black scales most
numerous.
4=Variable mix of white, orange, and sometimes gray scales; when present, the
gray scales may be concentrated along the dorsal midline and forming a solid
band.
Thorax:
78. Color of scales and hairs on meso- and metapleurons in the grooves where the femur
tucks in:
0=White scales and hairs only.
l=Light orange and black scales and hairs present.
2=Black scales and hairs present (some white scales and hairs may also be
present).
3=Black scales and hairs except orange scales above the coxae.
4=Predominately white with scattered black and orange scales.
@*A=1&2: Some individuals are state 1 while others are state 2.


176
Figures 166-168. Male genitalia capsules of Chlosyne sp. In dorsal aspect. 166: C. theona.
167: C. leanira. 168: C. lacinia. Fig. 169: Saccus of C. leanira (ventral aspect).


654
on the sides; inner lateral side with white scales and hairs (some black may be mixed in)
dorsally and at base and black scales and hairs ventrally and at tip. Vertex with a distinct
white centered patch reaching the posterior edge of the plate, only black scales elsewhere.
Sutures at the lateral edges of the vertex with black scales only. Tuft of white hairs on
the frontoclypeus anterior-lateral of each antennal base sparse, often appearing absent in
partially denuded specimens. Frontoclypeus with predominately black scales and hairs
with sparse white hairs usually present on the ventral part of the face. Patch of orange
scales absent at the edge of the inner anterior-lateral margin of the eye. Only black scales
present in the collar between the head and pronotum. Dorsally and laterally abdomen
solid black. Ventrally abdomen black with two lateral cream stripes (lateral to the
position of the parallel dark stripes in most Chlosyne, which are probably also present in
C. melanarge but indistinct against the largely black venter) occurring as disjunct
sections of variable size on a variable number of segments, midline black except for
small sections of cream on posterior segments in some specimens. Tibia and tarsi of
prothoracic legs with black scales dorsally and a variable mix of white and black scales
on the outer lateral and ventral sides. Femur of meso- and metathoracic legs covered
with black scales except orange at the distal most end. Tibia also covered with black
scales, except orange at the proximal most end. Tarsi with a variable mix of black and
orange scales with black most numerous. Meso- and metapleurons where the femur tucks
in with black scales and hairs. Thorax ventral to the wings with dense black scales and
hairs. Antennal shaft apparently lacking white patches of scales, with black scales and
more extensive areas devoid of scales relative to most Chlosyne, club with a white patch
on the outer lateral side in specimens where any scales are present in this area.


606
Locality: Mint Canyon. Holotype: Los Angeles County Museum (Miller and
Brown 1981).
=Melitaea leanira daviesi Wind, 1947. Pan-Pacific Entomologist. 23:171. Type
Locality: Strawberry, Tuolumne County, California. Holotype: California
Academy of Sciences (Miller and Brown 1981).
=ThessaIia leanira oregonenesis Bauer, 1975. The Butterflies of North America (by W.
E. Howe). Type Locality: will be designated as Oregon: Jackson County, Mt.
Ashland Loop Rd. (Austin and Smith 1998b). Lectotype: will be designated by J.
P. Donahue and deposited in the Los Angeles County Museum (Austin and Smith
1998b).
=Thessalia leanira elegans Priestaf and J. Emmel, 1998. Systematics of Western North
American Butterflies: 327-332. Type Locality: California: San Luis Obispo
County, Oso Flaco Sand Dunes, in the vicinity of Oso Flaco Lake. Holotype:
Natural History Museum of Los Angeles County.
=Thessalia leanira nebularum Austin and Smith, 1998. Systematics of Western North
American Butterflies: 335-336 and figures 7-8,15-16. Type Locality: California:
Santa Cruz County, Highland Way, 3 to 4 road miles southeast of Soquel-San
Jose Road. Holotype: Allyn Museum.
=Thessalia leanira austrima Austin and Smith, 1998. Systematics of Western North
American Butterflies: 339-340 and figures 45-46, 51-52. Type Locality: Mexico:
Baja California Norte, 53 road miles south of Catavina. Holotype: Allyn
Museum.
=Thessalia leanira basinensis Austin and Smith 1998. Systematics of Western North


REFERENCES
Ackery, P. R. 1988. Hostplants and classification: a review of nymphalid butterflies.
Biological Journal of the Linnean Society, 33: 95-203.
Austin, G. T. 1998. New subspecies ofNymphalidae (Lepidoptera) from Nevada
and Arizona). In Systemcitics of Western North American Butterflies, Thomas C.
Emmel, Editor. Mariposa Press. Gainesville, FL. Pp. 573-585.
Austin, G. T. and M. J. Smith. 1998a. Revision of the Thessalia leanira
complex (Lepidoptera: Nymphalidae: Melitaeinae): Thessalia leanira (C. & R.
Felder), with descriptions of four new subspecies. In Systematics of Western
North American Butterflies, Thomas C. Emmel, Editor. Mariposa Press.
Gainesville, FL. Pp. 333-358.
Austin, G. T. and M. J. Smith. 1998b. Revision of the Thessalia theona
complex (Lepidoptera: Nymphalidae: Melitaeinae). In Systematics of Western
North American Butterflies, Thomas C. Emmel, Editor. Mariposa Press.
Gainesville, FL. Pp. 359-396.
Bauer, D. L. 1959. A new species of Chlosyne (Nymphalidae) from Western Mexico. J.
of the Lepidopterists Society, 13: 165-169.
Bauer, D. L. 1960. Descriptions of two new Chlosyne (Nymphalidae) from Mexico,
with a discussion of related forms. J. of the Lepidopterists' Society, 14:148-154.
Bauer, D. L. 1961. Tribe Melitaeinae. In How to Know the Butterflies by Paul R.
Ehrlich and Anne H. Ehrlich. W. C. Brown Co., Dubuque, Iowa.
Bauer, D. L. 1975. Tribe Melitaeini. In The Butterflies of North America by W. H.
Howe. Doubleday & Company, Garden City, New York.
Behr, H. H. 1863. On California Lepidoptera No. 3. Proceedings of the California
Academy of Natural Sciences, 3: 84-93.
Bell, G. 1997. The Basics of Selection. Chapman and Hall, New York.
Brown, F. M. 1966. The types of Nymphalid butterflies described by William
Henry Edwards-Part II, Melitaeinae. Transactions of the American
Entomological Society, 92: 357-468.
790


615
the Oklahoma record is either an unusual stray or transport (it is, however, in fresh
condition) or of suspicious authenticity.
Chlosyne leanirafulvia W. H. Edwards
Melitaeo fulvia W. H. Edwards, 1879. Trans. Amer. Ent. Soc. 3:191. Type Locality:
Texas, restricted to Archer County by Brown (1966). Lectotype: Carnegie
Museum (Brown 1966).
=Thessalia fulvia pariaensis Smith and Brock, 1988. Bulletin of the Allyn Museum. No.
118: 10-12. Type Locality: Utah, Kane County, the Cockscomb Ridge, ca. 61
road km east of Kanab. Holotype: Los Angeles County Museum.
=Thessalia fulvia coronaclo Smith and Brock, 1988. Bulletin of the Allyn Museum. No.
118: 7-10. Type Locality: Arizona: Pima County, summit of Redington Pass,
20km ENE of Tucson, Santa Catalina Mts. Holotype: Los Angeles County
Museum.
Diagnosis: I have seen many specimens of C. leanira fulvia confused with C. leanira
leanira and C. leanira cyneas. C. leanira fulvia may be consistently separated from C.
leanira leanira by the absence of orange scales on the labial palpi. However, the wing
pattern and wing shape are identical to many specimens of the alma (mc\. flavoclorsails
and basinensis phenotypes) phenotype of C. leanira leanira. Populations with females of
subspecies fulvia resembling the alma phenotype have a marked sexual dimorphism with
males distinctly darker than females, while males tend to be at most slightly darker than


431
Strict
96
CI=.825
RI=.916
RC=. 756
i2
69
59
62
54
56
55
85
75
89
72
70
63
53
95
85
67
95
69
59.
99
91
99
99,
100
- C. nycteis
- C. gorgone
- C. harrissii
- C. kendallorum
- C. hoffmanni
- C. palla
- C. gabbii
- C. acastus
- C. w. damoetas
- C. w. whitneyi
- C. definita
- T. ezra
- T. theona
- T. perlula
- T. chinatiensis
- T. I. leanira
- T. I. fulvia
- T. cyneas
- T. cynisca
- C. e. pardelina
C. e. endeis
- C. melitaeoide;
- C. marina
- C. e. poecile
- C. e. erodyle
- C. melanarge
C. eumeda
- C. cal ¡fornica
- C. lacinia
C. ehrenbergi
- C. hippodrome
C. narva
C. g. gaudealis
C. g. wellingi
C. janais
C. marianna
C. rosita rosita
C. montana
C. rosita browni
C. riobalensis
C. mazarum
A. perezi
A. pantoni
A. tulita
T. eleda
D. dymas
T. coracara
T. anomalus
M. elva
H. fasciatus
H. miriam
A. pelops
A. proclea
P. minuta
Figure 335: Strict consensus tree from a heuristic search of successively weighted pattern characters from
the Chosyniti/Poladryiti data matrix with all genitalic characters (char 1-52) deleted.


579
Rica/Campeche specimens. In all C. theona specimens I have examined the dorsal
forewing median band is distinctly bordered by black basally in cells CuA2 and CuAl,
while in C. perlula there is no black or only diffuse black between the median band
patches and orange basal area in these cells. The same is true for the periphery of the
pale cream spot in the dorsal forewing discal cell. The basal areas on the dorsal hindwing
and forewing are always predominately orange with only diffuse black scaling in C.
perlula, and while this is the case in a portion of C. theona from Costa Rica and
Campeche none are as extreme as C. perlula. The dorsal hindwing postmedian band is
always bordered by black basally in C. theona, but this is not the case in one of four C.
perlula I have examined. The ventral wing surfaces are variable in both C. perlula and
Costa Rican C. theona, however one consistent difference appears to be sparse black
scaling around the periphery of the cream patch at the distal end of the forewing discal
cell in C. perlula versus a distinct black border in C. theona. Also, C. perlula tends to
have a broader black band bisecting the ventral forewing median band between CuA2
and M3 relative to C. theona, and with the exception of the Costa Rica and Campeche
phenotypes C. theona lacks a ventral forewing black band bisecting the median band
altogether.
Further Description: It must be stressed that the following description is based on the
examination of only four specimens, none of which had all body parts and appendages
intact but were in fresh condition in terms of wing pattern. Austin and Smith (1998a)
provide black and white photographs of a series of six specimens.
The pattern on the body and appendages is like that described for C. chinatiensis
above with a few exceptions. C. perlula lacks any orange scales on the vertex. The


489
the ventral hindwing, sections span the width of cells or not but are not aligned. Dorsal
hindwing margin distal to the position of the submarginal band black without orange
scaling, ventrally with a diffuse orange band in the equivalent position on both the
forewing and hindwing. Ventral hindwing submarginal band distinct against the
background. Dorsal forewing median band white and dorsal hindwing median band
absent. Dorsal hindwing between distal parafocal element and submarginal band with
diffuse crescents against a black background, sometimes so reduced that they are difficult
to see without magnification, ventrally in the equivalent position the crescents are sharp
and well defined, although quite narrow except for cell M3. Characteristics of the ventral
hindwing bands of the basal and central symmetry system and the bands between them
are as described in M. eleda (above). Ventral hindwing lacking heavy black scaling
along the veins. Background color basal to the basal parafocal elements cream,
appearing as more of an off white in more worn specimens. Ventral hindwing
background color between the proximal band of the central symmetry system and the
distal band of the basal symmetry system concolorous with the remaining background
basal to the basal parafocal elements.
Variation: There appears to be very little variation other than small differences in the
size of wing pattern elements and a tendency of females to be larger than males.
Range: I have seen specimens only from the adjacent Mexican states of Guerrero
(Iguala, Naranjo, and "Mexeala") and Morelos (10 miles south of Cuemevaca).
Collection data from the NMNH and AM are presented in Figure 12.
Species Delimitation: See M. anomalus below.


729
Diagnosis: The only differences which I find between subspecies mazarum and
riobalensis are that mazarum has a pale cream/yellow basal band on the dorsal hindwing,
and the ventral hindwing has a cream colored basal area of comparable extent to
subspecies montana.
Further Description: It is not possible to generalize about the range of variation as only
two male specimens were examined, but all characters except for those noted above are
in the range of variation exhibited by C. rosita riobalensis. The holotype has all of the
light markings present and conspicuous on the dorsal forewing with the anterior median
band patches expanded, while the other specimen has most forewing light markings other
than the anterior median band patches small or absent dorsally. The ventral hindwing
median band is distinctly separated from the light basal area by an area of black in both
specimens.
Distribution of Specimens Examined: Other than the holotype (locality listed above
above), I have seen only one specimen of this taxon, labeled: Mexico: Oaxaca: Hwy 135,
about 10 miles S ofCuigatlan, 15 July 1988, John Kemner (AM,lc?). I suspect that
Scott's (1986) reports of Chlosyne rosita mazarum from Arizona and W. Mexico are
actually misdeterminations of C. rosita montana.


691
ehrenbergi are mentioned under C. ehrenbergi. C. hippodrome differs from C. gaudealis
and C. narva by having a strongly curved inner valve process in ventral view (Figure 151
versus Figures 152-153). The shape of the inner valve process of C. hippodrome in
ventral view separates it from all other Chlosyniti. Also, the forks of the saccus are
distinctly farther apart in C. hippodrome relative to either C. gaudealis or C. narva.
Further Description: Black and white but no orange scales and hairs on the labial palpi.
Dorsally palpi with black scales and hairs only; outer lateral side white ventrally-center
and at base while black dorsally and at tip, black hairs around the edge with some white
ones mixed in on the ventral side; ventrally with base white and black hairs and scales in
the center and throughout the terminal segment, with the white lateral bands extending
down along the sides; inner lateral side with white scales and hairs (some black may be
mixed in) dorsally and at base and black scales and hairs ventrally and at tip (some
specimens may have sparse white scales mixed in at the tip). Vertex with a distinct white
centered patch reaching the posterior edge of the plate, only black scales elsewhere.
Sutures at the lateral edges of the vertex with black scales only. Tuft of white hairs on
the frontoclypeus anterior-lateral of each antennal base well developed. Frontoclypeus
with black scales and hairs as background, also with white patches of scales touching the
eyes ventro-laterally of the antennal bases and a white patch centered on the ventral side
of the face. Patch of orange scales absent at the edge of the inner anterior-lateral margin
of the eye. Black and sometimes white scales present in the collar between the head and
pronotum. Abdomen covered with black scales throughout on the dorsal and lateral
sides, ventrally predominately black but with sparse and diffuse cream scaling along the
midline and to the sides of the midline (in the equivalent position to where solid bands of


387
Another interesting observation to draw from this case study is that there are no
unambiguous pattern character state changes supporting the nodes for Poladryiti,
Chlosyniti, and Microtia, only one for Microtia + Chlosyne (a nonhomoplastic character),
one for Higginsius + Atlantea (a homoplastic character), and of the five unambiguous
pattern character state changes at the Chlosyne node, three are for homoplastic characters
and the remaining two do not unambiguously represent universal synapomorphies
(character 97 occurs as two states within Chlosyne and character 109 cannot be scored for
all taxa). While genitalic character support for these nodes is robust, there are no wing
pattern characters which by themselves support them. On the other hand, for many nodes
where there are no or few unambiguous genitalic character state changes and/or universal
synapomorphies from genitalic characters, support from pattern characters is robust. As
noted within the Chlosynit and Poladryiti, on average (with exceptions) genitalic
characters appear to be more stable over evolutionary time relative to wing pattern
characters. This is evidenced by the observation that of the 97 character states requiring
more than the minimum number of steps on the most parsimonious tree (based on the
distinct state model, see Table 11), only nine are from genitalic characters while 88 are
from pattern characters, including six versus 69 for states requiring one extra step, two
versus fourteen for states requiring two extra steps, and one versus five for states
requiring three extra steps, for genitalic versus pattern characters, respectively. Of
course, since 64.9% of the total characters are pattern characters the comparison is not a
direct one, but the differences are nonetheless obvious. Furthermore, most genitalic
characters have only two or three character states in the data set, while many more states
are common for pattern characters. At least in this case study, a general pattern appears


763
APPENDIX A
PLATES OF ADULT CHLOSYNITI & POLADRYITI
Plate A: Poladryiti (Adults) Plate D: Chlosyniti (Adults)
1 Atlantea tulita
if
D
Puerto Rico. Quebradilla, Puerte Blanco
1 Microtia coracara
9
V
Mexico. Guerrero. Iguala
2 Atlantea tulita
if V
Same as above
2 Microtia coracara
9
D
Mexico. Morelos. Hualintlan
3 Atlantea tulita
9
D
Puerto Rico
3 Microtia coracara
9
V
Same as above
4 Atlantea tulita
9
V
Puerto Rico
4 Microtia anomalus
tf
D
Mexico, Michoacan, Coahuavana
5 Atlantea cryptadia
if
D
HOLOTYPE: Haiti. rOuest, BoutiHiers Rd.
5 Microtia anomalus
V
Same as above
6 Atlantea cryptadia
if
V
Same as above
6 Chlosyne harrisii
9
D
Canida. Manitoba. Miami
7 Atlantea pantoni
if
D
Jamaica. Crown Lands nr. Trov
7 Chlosyne harrisii
9
D
Canada. Quebec. Temiscouata Co
8 Atlantea pantoni
if
V
Same as above.
8 Chlosyne harrisii
9
V
Same as above
9 Atlantea pantoni
9
D
Jamaica. Crown Lands nr. Trov
9 Chlosyne harrisii
9
D
USA. New Hampshire. Coos Co.
10 Atlantea pantoni
9
V
Same as above
10 Chlosyne harrisii
cf
D
USA. Pennsylvania. Clinton Co.
11 Atlantea pantoni
9
D
11 Chlosyne harrisii
tf
V
Same as above
12 Atlantea pantoni
9
V
12 Chlosyne harrisii
9
D
Canada. New Brunswick, Edmundston
13 Atlantea perezi
if
D
Cuba. Holguin
13 Chlosyne harrisii
9
V
Same as above
14 Atlantea perezi
if
V
Same as above
14 Chlosyne harrisii
9
D
Pennsylvania, Centre Co
15 Atlantea perezi
9
D
No Data
15 Chlosyne harrisii
9
V
Same as above
16 Atlantea perezi
9
V
Same as above.
16 Chlosyne kendallorum
cf
M
Mexico. Nuevo Leon. 40 km WSW of Cola
17 Hippinsius fasciatus
if
D
Peru. Cuzco, Torontoy. 7000 ft
17 Chlosyne kendallorum
tf
V
Same as above
18 Hippinsius fasciatus
if
V
Same as above
18 Chlosyne kendallorum
9
M
Mexico. Nuevo Leon, Hwy 60 59 km WSW
19 Hippinsius miriam
if D
Ecuador. Loja
19 Chlosyne kendallorum
9
V
Same as above
20 Hippinsius miriam
if V
Same as above.
20 Chlosyne nycteis
tf
D
Manitoba. Pine Ridge
21 Hippinsius miriam
9
D
Ecuador. Azuav. Q Corazn Urca nr Azogues
21 Chlosyne nycteis
D
USA, Virginia. Montgomery Co
22 Hippinsius miriam
9
V
Same as above
22 Chlosyne nycteis
tf
V
Same as above
23 Poladryas minuta
if
D
USA. Arizona, Coconino Co.
23 Chlosyne nycteis
tf
D
USA. Pennsylvania, Centre Co
24 Poladryas minuta
if V
Same as above.
24 Chlosyne nycteis
tf
V
Same as above
25 Poladryas minuta
USA. California. Tulare Co.
25 Chlosyne nycteis
tf
D
USA. Georgia. Cobb Co.
26 Poladryas minuta
if
V
Same as above
26 Chlosyne nycteis
tf
D
USA, Pennsylvania. Marietta
27 Poladryas minuta
if
D
USA. Texas
27 Chlosyne nycteis
tf
V
Same as above
28 Poladryas minuta
if V
Same as above
28 Chlosyne nycteis
tf
D
USA. Colorado. Park Co.
29 Chlosyne nycteis
tf
V
Same as above
Plate B: Poladryiti & Chlosyniti (Adults) Plate E: Chlosyniti (Adults)
1 Poladryas minuta
tf D
USA. New Mexico
1 Chlosyne nycteis
tf
V
Canada. Manitoba. Pine Ridge
2 Poladryas minuta
cf V
Same as above
2 Chlosyne nycteis
tf
V
USA. Colorado. Boulder Co
3 Poladryas minuta
USA. Arizona. Santa Cruz Co.
3 Chlosyne nycteis
tf
D
USA. New Mexico. Jemez Mts
4 Poladryas minuta
Same as above
4 Chlosyne nycteis
9
D
USA, Texas. Hamilton Co.
5 Poladryas minuta
9 D
USA. Colorado. Sedalia
5 Chlosyne nycteis
9
D
USA. Virginia. Fairfax Co
6 Poladryas minuta
9 V
Same as above
6 Chlosyne nycteis
9
V
Same as above
7 Poladryas minuta
9 D
USA, Colorado. Gilpin Co
7 Chlosyne nycteis
9
V
Same as No. 4
8 Poladryas minuta
9 V
Same as above
8 Chlosyne nycteis
9
D
USA. Kansas. Johnson Co
9 Antilles proclea
Jamaica
9 Chlosyne nycteis
9
V
Same as above
10 Antilles proclea
Same as above
10 Chlosyne nycteis
9
V
No Data
11 Antillea proclea
9 D
Jamaica
11 Chlosyne porpone
tf
D
USA, Kansas. Johnson Co
12 Antillea proclea
9 V
Same as above
12 Chlosyne porpone
tf
D
USA. Mssoun, Jackson Co
13 Antillea petops
tf D
Dominican Republic. Santiago
13 Chlosyne porpone
tf
D
USA. Iowa. Ames
14 Antillea petops
tf V
Same as above
14 Chlosyne porpone
9
D
USA, Georgia. Dekalb Co
15 Antillea petops
9 D
Dominican Republic, Samana. San Domingo
15 Chlosyne porpone
tf
D
USA, Missouri. Benton Co.
16 Antillea petops
9 V
Same as above
16 Chlosyne porpone
f
V
Same as above
17 Microtia eleda
tf D
Ex egg Female. USA. Texas. Hidalgo Co
17 Chlosyne porpone
tf
V
USA. Colorado. Boulder Co
18 Microtia eleda
tf V
Same as above
18 Chlosyne porpone
if
V
USA. Missouri. Cass City
19 Microtia eleda
tf D
USA. Arizona. Cochise Co.
19 Chlosyne porpone
tf
V
USA, Kansas. Johnson Co.
20 Microtia eleda
tf V
Same as above
20 Chlosyne porpone
9
D
USA. Colorado. Turkey Creek
21 Microtia eleda
tf V
USA. Arizona. Cochise Co.
21 Chlosyne porpone
9
V
Same as above
22 Microtia eleda
tf V
USA. Arizona. Cochise Co.
22 Chlosyne porpone
9
V
USA. Oklahoma. Norman
23 Microtia eleda
tf D
Mexico. Oaxaca, Monte Alban
23 Chlosyne porpone
9
D
USA. Georgia. DokaJb Co.
24 Microtia eleda
tf V
Same as above
24 Chlosyne porpone
9
D
USA, Kansas, Barber
25 Microtia eleda
9 D
USA. Arizona. Santa Rita Mts
25 Chlosyne hoffmanni
tf
D
USA, Washington. Yakima Co
26 Microtia eleda
9 V
Same as above
26 Chlosyne hoffmanni
tf
D
Same as above
27 Microtia eleda
9 D
Ex egg Female. USA. Texas. Hidalgo Co.
27 Chlosyne hoffmanni
tf
D
USA. Washington. Klickitat Co
28 Microtia eleda
9 V
Same as above
28 Chlosyne hoffmanni
tf
D
USA, Washington. Yakima Co
29 Microtia eleda
9 D
USA. Arizona. Cochise Co.
Plate C: Chlosyniti (Adults)
1 Microtia eleda 9 D
2 Microtia eleda 9 V
3 Microtia elva ¡f D
4 Microtia elva & D
5 Microtia elva 6 Microtia elva 9 V
7 Microtia elva 9 D
8 Microtia elva 9 D
9 Microtia elva 9 V
10 Microtia dymas 9 D
11 Microtia dymas 9 V
12 Microtia dymas 9 D
13 Microtia dymas 14 Microtia dymas tr D
15 Microtia dymas tf D
16 Microtia dymas if D
17 Microtia dymas if D
18 Microtia dymas if V
19 Microtia dymas if V
20 Microtia dymas if D
21 Microtia dymas 9 D
22 Microtia dymas 9 V
23 Microtia dymas 9 D
24 Microtia coracara 25 Microtia coracara tf V
26 Microtia coracara 9 D
Mexico. Hidalgo. Ixmiquilpan
Same as above
Mexico. Tamaulipas, 49 km S of Cd. Victoria
Mexico. Sinaloa, 32.2 mi. NE Villa Union
Mexico. Tamaulipas. 2 mi SW of Ciudad Manle
Mexico. Veracruz. Rinconada
Same as above
Mexico. Guerrero. Acapulco
Same as above
USA. California. San Diego Co.
Same as above
USA, California. Riverside Co.
USA. Texas. San Patricio Co.
USA. Texas. Culberson Co.
USA. California, Riverside Co.
USA. California. San Diego Co.
USA. Arizona. Pima Co
USA. California. San Diego Co.
USA. Texas. San Patricio Co,
USA. California, San Diego Co
USA, Arizona. Pima Co.
USA. Texas. San Patricio Co
USA. Texas. San Patricio Co.
Mexico. Guerrero. Mexeala
Same as above
Mexico. Guerrero. Iguala
Plate F: Chlosyniti (Adults)
1 Chlosyne hoffmanni if V
2 Chlosyne hoffmanni tf D
3 Chlosyne hoffmanni if V
4 Chlosyne hoffmanni tf D
5 Chlosyne hoffmanni tf D
6 Chlosyne hoffmanni if D
7 Chlosyne hoffmanni if D
8 Chlosyne hoffmanni tf V
9 Chlosyne hoffmanni if V
10 Chlosyne hoffmanni 9 D
11 Chlosyne hoffmanni ¡f D
12 Chlosyne hoffmanni if V
13 Chlosyne hoffmanni 9 D
14 Chlosyne hoffmanni 9 V
15 Chlosyne palla if D
16 Chlosyne palla if V
17 Chlosyne palla if D
18 Chlosyne palla if V
19 Chlosyne palla if D
20 Chlosyne palla if D
21 Chlosyne palla 22 Chlosyne palla 9 D
23 Chlosyne palla 9 V
USA. Washington. Yakima Co
USA. Washington. Yakima Co
Same as above
USA. Washington, Yakima Co.
USA. California. Sierra Co
USA. California. Sierra Co.
USA. California. Sierra Co
Same as above
Aberration; USA. Washington. Okanogan Co.
USA. Washington, Yakima Co
USA. Washington, Chelan Co
Same as above
USA. California. Sierra Co
USA. California. Sierra Co.
USA. California. El Dorado Co
Same as above
USA. California
Same as above
USA. Washington. Okanogan Co
USA. California. El Dorado Co
Same as above
USA. California. Plumas Co
USA. California. Plumas Co


469
orientation and distribution of setae compared with any Chlosyniti). Ventral surface of
the juxta with a plateau which is either diamond-shaped (Figures 108 & 134-155),
triangular (Figures 110 & 122) or somewhat hour-glass shaped (Figure 113). Ventral
surface of the abdomen with two longitudinal dark stripes (some other derived states
occur within Chlosyne in taxa with darker abdomens, but this series of ventral abdominal
pattern character states is unique among all Melitaeini examined).
Microtia Bates
Microtia Bates, H. W., 1864. Ent. mon. Mag. 1:83. Type species: Microtia elva Bates.
=Texola Higgins, 1959. Lepid. News 12: 161. Type species: Eresia eleda Hewitson.
=Dymasia Higgins, 1960. Trans. Royal Ent. Soc. London. 112: 455-456. Type species:
Melitaea dymas W. H. Edwards.
Synapomorphies from binary characters: A raised posteriorly orientated hollow and
compressed ridge on the lamella antevaginallis forming roughly a half circle around the
ventral genital opening (Figures 279-282).
A possible additional synapomorphy from the immature stages is the scolus on A9
reduced to a sclerotized plate as reported by Harvey (1992). Harvey (1992) reported
having examined the genera Texola, Dymasia, Microtia, Chlosyne, Thessalia and
Poladryas (plus genera outside the Poladryiti and Chlosyniti), and of these genera only
Texola, Microtia, and Dymasia have the sclerotized plate. Which species in these genera
were examined is not reported, however it can be inferred that dymas and elva were


338
polytomies, including unambiguous ones. Refer to Figures 312 and 316 for the actual
strict consensus trees. Status of individual characters in these two analyses, including the
number of steps required for each character on the most parsimonious tree, character
consistency indices, retention indices, and rescaled consistency indices, are presented in
Tables 7 and 8 for the regular PAUP analysis and the distinct state model, respectively.
In addition to the six different parsimony analyses with a cumulative out group
(successively and equally weighting characters for the three different models for
multistate taxa), I conducted the same six analyses with Poladryiti designated as the out
group. Furthermore, I conducted the same six analyses with Chlosyniti only, with
Antillea and Microtia (including Texola and Dymasia) designated as the out group and
Chlosyne (including Thessalia and Charidryas) designated as the in group. All these
analyses produce the same strict consensus trees for their respective taxa and model for
polymorphic characters as analyses based on the cumulative out group.
Variation in Tree Statistics with Different Numbers of Taxa in the Analysis
Table 9 includes values obtained for tree consistency index, retention index, and
rescaled consistency index with different numbers of taxa in the analysis for equally and
successively weighted characters. Node numbers, indicating the clade used in each
analysis, are based on the nodes designated in Figure 312. The relationship between
numbers of taxa in the analysis and tree statistics is presented graphically in Figure 328.
All tree statistics exhibit a general pattern of decreasing as larger monophyletic
groups are included in the analysis; however, the extent of the decrease varied for
different statistics. Also, in this case study the tree statistics initially underwent a steep
decrease once they began to drop below 1.0, but this was followed by a slight decrease or


530
California. Lectotype: National Museum of Natural History (J. Emmel et al.
1998d).
=Melitaeapalla eremiata Wright, 1905. Butt. West Coast: 160. Type Locality: Central
California, fixed to San Rafael, Marin County, California (J. Emmel et al. 1998a).
Lectotype: California Academy of Sciences (Miller and Brown 1981).
=Melitaeaflavula Barnes and McDunnough, 1918. Contrib. Lep. N. Amer. 4:73. Type
Locality: Colorado, restricted to Genwood Springs, Colorado. Holotype: NMNH
[examined].
=Melitaea sterope W. H. Edwards, 1870. Trans. Am. ent. Soc. 3:190. Type Locality:
Oregon, restricted to Tygh Valley, Wasco County, Oregon, by Bauer (1975).
Lectotype: Carnegie Museum (Miller and Brown 1981).
=Melitaea calydon Holland, 1931. Butterfly Book, revised ed.: 125. Type Locality:
Turkey Creek Junction, Jefferson County, Colorado. Syntypes: Carnegie
Museum (Miller and Brown 1981).
-Melitaea hewesi Leussler, 1931. Ent. News 42:12. Type Locality: Tygh, Oregon.
Holotype: "Probably at Ohio St. University" (Miller and Brown 1981)
=?Melitaea sterope form hopfingeri Gunder, 1934. Caad. Ent. 66:129. Type Locality:
Brewster, Washington.
=Chlosyne (Chariclryas) palla altas ierra Emmel, Emmel, and Mattoon, 1998.
Systematics of Western North American Butterflies: 140-141 & 155 (Figures 5-
8). Type Locality: California: El Dorado County, Fallen Leaf Lake, elevation ca.
6400. Holotype: Natural History Museum of Los Angeles County.
=Chlosyne (Charidryas) palla australomontana Emmel, Emmel, and Mattoon, 1998.


214
relationships among taxa (i.e. grouping by synapomorphies and use of parsimony).
Rather, Higgins (1960 and 1981) classifications were based on the criteria of perceived
similarities and differences, including grouping by symplesiomorphies and splitting by
autapomorphies. This observation is exemplified by the following quotes: regarding a
monotypic concept of Microtia: a insect which shows several divergent features and
remains the only member of the genus (Higgins 1960); regarding placing dymas in a
separate genus from elacla: these differences are so marked that I have not found it
possible to write a generic synopsis to include both dymas and elada" (Higgins 1960);
and regarding Charidryas: The specialised characters do not appear to be sufficiently
marked to justify generic separation (Higgins 1960). Classifications based on ones
opinions of the degree of similarity or divergence are inherently unstable, as such
concepts and alternatives based on opposing views of the degree of similarity or
difference represent arbitrary opinions rather than testable hypotheses. While there is
only one phylogeny, there are numerous ways that different assemblages of taxa can be
viewed as similar or different, and the scientific method cannot be used to choose among
them. Furthermore, the groups whose recognition is based on perceived similarity or
degree of divergence will often be artificial, because dissimilarity can be due to
autapomophies and similarity can be due to symplesiomorphies, and neither of these
provide evidence of natural groups. Consequently, the validity of the genera used in
Higgins (1960 and 1981) classification in a natural classification scheme was in need of
testing.
In addition to deriving a phylogenetic hypothesis for Chlosyniti and Poladryiti, I
used this data set as a case study to investigate some issues of broader application to


539
Diagnosis: C. whitneyi whitneyi is more difficult to separate from C. cicastus and C. palla
than is the case for C. whitneyi clamoetas, and I elect to include separate diagnoses for the
two allopatric subspecies which I recognize (whitneyi Behr and clamoetas Skinner)
without a combined diagnosis.
Further Description: The pattern of the body and appendages is like that described for C.
hoffmanni with the exception of the frontoclypeus, femur of the meso- and metathoracic
legs, and part of the meso- and metathoracic pleurons where the femur tucks in. The
femur pattern varies between subspecies (see below) but some black scales are present
(not found in a some denuded individuals) in contrast to other taxa in the C. hoffmanni
clade. Some (but not all) specimens lack orange scales on the frontoclypeus (more
common in subspecies damoetas than subspecies whitneyi), which is never the case in
other members of the C. hoffmanni clade. The meso- and metathoracic pleurons where
the femur tucks in may or may not have pale orange scales in addition to white ones (with
some black scales underneath) as in C. acastus and C. gahhii (C. hoffmanni and C. palla
apparently always have pale orange scales present).
Both subspecies whitneyi and clamoetas include all of the wing pattern elements
described for C. hoffmanni above. The dorsal light markings are pale orange with little or
no contrast between the median and postmedian bands or other light markings. The
hindwing is always diffused with black in the basal area, and the borders of the symmetry
system and parafocal element black markings have a somewhat more diffused appearance
relative to similar taxa where the distinction between black and light markings is more
contrasted. The hindwing background color is white without a cream tint.


CHAPTER 5
SUMMARY AND CONCLUDING REMARKS
This work presents a phylogenetic hypothesis for a major group of Nymphalid
butterflies based on two extensive morphological data sets, with the relationships among
most members of the Melitaeini resolved (excluding only the relationships among taxa
within the Phycioditi and Melitaeiti clades). The proposed phylogeny is supplemented by
detailed descriptions of characters and character states and how they were coded, detailed
camera lucida drawings illustrating all genitalic character states, and data matrices
showing how all taxa were coded for each character.
The second chapter revises the Melitaeini at the subtribal level, identifying the six
primary clades comprising the trib. These clades are the subtribes Euphydryiti,
Melitaeiti, Phycioditi, Gnathotrichiti, Poladryiti, and Chlosyniti. The decision to
recognize these six clades as subtribes was based on the criteria of creating a natural and
stable higher classification. Each subtribe is described in detail, including evidence
supporting subtribal monophyly and the relationships among subtribes. Three of the
subtribes (Euphydryiti, Gnathotrichiti, and Poladryiti) are revised at the generic level,
with monophyletic genera comprising these tribes identified and described. The
information learned from this study provides the foundation of the study in the following
chapter. The taxa comprising the lineage including the genus Chlosyne were identified,
along with the relationships of this lineage to the remaining Melitaeini, providing the
757


127
4 Males: Valvae orientated much more ventrally than posteriorly (Fig. 76), posterior
inner wall of valvae not so extended, posterior end of valve terminates in either a short,
curved tooth (Fig. 74) OR a tubular extension with a flared, flatter end.
Females: Sclerotized plate on the ventral corpus bursae curved posteriorly at its anterior
end (Figs. 269-272), a distinct sclerotized tube (ostium bursae) extending distal of the
plates of lamella postvaginallis and lamella antevaginallis (Figs. 269-270 & 272), a
membranous pouch with many longitudinal wrinkles on the ventral surface originating at
the edge of the lamella antevaginallis and extending posteriorly over the lamella
antevaginallis and genital opening (Fig. 269).
Distribution: Nearctic and Neotropical Phycioditi
5. Males: The teeth bearing plate on the vesica with the teeth in one row along an arch
(Figs. 105 & 211-212), entire posterior end of phallus extending distally as a broad,
flattened, non-hollow extension with the sides convex at the base (Figs. 1 & 104),
tegumen in the form of a narrow sclerotized bridge lacking paired, hollow, posterior
projections.
Females: Lamella postvaginallis and lamella antevaginallis project at acute angles
forming a partial open pouch around the genital opening, plate on ventral corpus bursae
extending anteriorly as paired projections Chlosyniti
5. Males: The teeth bearing plate on the vesica with the teeth not confined to one row
(Figs. 78 & 83-84), midline of the posterior end of phallus extending distally in an
elongate, tapering triangular extension with a hollow tube at its distal end (Figs. 83,88,92
& 97-98), tegumen expanded posteriorly and with paired hollow posterior projections
(Figs. 80-81,86,90 & 95).


129
Table 1: Data Matrix for a Phylogenetic Analysis of the Melitaeini
Characters
Taxa
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19 20
Chlosynel
2
1
1,2
2
0
0
0
0
0
0
?
0
1
1
1
1
1
2
0 1
Chlosyne2
2
1
1,2
2
0
0
0
0
0
0
?
0
1
1
1
1
1
2
0 1
"Texola elada
2
1
2
2
0
0
0
0
IF
0
?
0
1
1
2
1
1
2
0 1
*Microtla elva
2
1
2
2
0
0
0
0
0
0
?
0
1
1
2
1
1
2
0 1
"Dymasia dymas
2
1
2
2
0
0
0
0
0
0
?
0
1
1
2
1
1
2
0 1
Texola corocara
2
1
2
2
0
0
0
0
0
0
?
FF
1
1
2
1
1
2
0 1
Texola anomalus
2
1
2
2
0
0
0
0
0
0
?
0
1
1
2
1
1
2
0 1
"Antillea pelops
2
1
1
2
0
0
0
0
0
0
?
0
1
1
2
1
2
1
0 1
Antillea proclea
2
1
1
2
0
0
0
0
0
0
?
0
1
1
2
1
2
1
0 1
*Polydryas minuta
2
1
1
2
0
0
0
0
0
0
?
0
1
1
2
1
3
3
0 1
*Higginsius fasciatus
2
1
1
3
0
0
0
0
0
0
?
0
1
1
2
1
1
2
0 1
Hlgginsius mlriam
2
1
1
3
0
0
0
0
0
0
?
0
1
1
2
1
1
1
0 1
Atlantea tulita
2
0
1
5
0
0
0
0
0
0
?
0
1
1
2
2
?
1
0 1
Atlantea pantoni
2
0
1
5
0
0
0
0
0
0
?
0
1
1
2
2
?
1
0 1
"Mellicta athalia
2
1
1
4
0
0
0
0
0
0
?
2
1
1
2
3
?
1
0 1
Melllcta brltomartis
2
1
1
4
0
0
0
0
0
0
?
nr
1
1
2
3
?
1
0 1
Melllcta aurella
2
1
1
4
FT
0
0
0
0
0
?
1
1
1
2
3
?
1
0 1
Melllcta varia
2
1
1
4
0
0
0
0
0
0
?
1
1
1
2
4
?
1
0 1
Melllcta parthenoides
2
1
1
4
0
0
0
0
0
0
?
1
1
1
2
5
?
1 0 1
Mellicta asteria
2
1
1
4
0
0
0
0
0
0
?
2
1
1
2
4
?
1
0 1
Mellicta alatauica
2
1
1
4
0
0
0
0
0
0
?
0
1
1
IF
3
?
1
0 1
"Cinclidia phoebe
2
1
1
2
0
0
0
0
0
0
?
0
1
1?
?
?
?
1
1 1
Clnclldia scotosla
2
1
1
2
0
0
F
0
0
0
?
0
1
1?
?
?
?
1
1 1
"Melitaea cinxia
2
1
1
2
0
0
0
0
0
0
?
IF
1
0
0
0
?
4
0 1
Melitaea diamina
2
1
12
0
0
0
0
0
0
?
0
1
0
F
0
?
4
0 1
"Didymaeformia didyma
2
1
1
3
0
0
0
0
0
IF
?
0
1
0
0
0
0
5
0 1
Didymaeformia trivia
2
1
1
6
0
0
0
0
0
0
?
0
1
0
0
0
0
5
0 1
"Phyciodes tharos
2
1
1
2
0
0
0
0
0
0
?
0
1
0
0
0
0
1
0 1
Phyciodes batesi
2
1
1
2
0
0
0
0
0
0
?
0
1
0
^F
0
0
1
0 1
Phyciodes phaon
2
1
1
2
0
0
0
0
0
0
?
0
1
0
0
0
0
1
0 1
"Phystis simois
2
1
1
2
0
0
0
0
0
0
?
0
1
0
0
0
0
1
0 1
*Anthanassa texana
2
1
1
2
0
0
0
0
0
IT
?
0
1
0
0
0
0
1
0 1
Anthanassa tulcis
2
1
1
2
0
0
0
0
0 0
?
0
1
0
0
0
0
1
0 1
*Telenassa teletusa
2
1
1 2
0
0
fir
0
0
0
?
0
1
0
0
0
FT
1
0 1
*Castilia castilla
2
1
1
2
0
0
0
0
0
0
?
nr
1
0
0
0
0
1
0 1
*Dagon catula
2
1
1
2
0
0
0
0
0
0
?
0
1
0
0
0
0
1
0 1
*Ortilia liriope
2
1
1
2
0
0
0
0
0
0
?
0
1
0
0
0
0
1
FTT
Ortilia ithra
2
1
1
7
0
0
0
0
0
0
?
0
1
0
0
0
0
1
0 1
*Eresia eunice
2
1
1
2
0
0
0
0
0
0
?
0
1
0
0
0
0
1
0 1
Eresia frisia
2
1
1
2
0
0
0
0
0
0
?
0
1
0
0
0
0
1
~o\T
*Janatella leucodesma
F
1
1
2
0
0
0
0
0
0
?
0
1
0
0
0
0
1
0 1
Janatella fellula
F
1
1
2
0
0
o
IT
0
0
?
0
1
0
0
0
0
1
0 1
Mazia amaznica
2
1
1
2
0
o
0
0
0
0
?
0
1
0
0
0
0
1
0 1
*Tegosa claudina
2
1
1
2
0
0
~o1
0
0
0
?
0
1
0
0
0
0
1
0 1
Tisona saladellensis
2
1
1
2
0
0
0
0
0
0
?
0
1
0
0
0
0
1
0 1
*Gnathotriche exclamationis
2
2
1
2
0
0
0
0
0
0
?
0
1
0
0
0
0
1
0 1
Gnathotriche sodialis
2
2
1
2
0
0
0
0
0
0
?
0
1
0
0
0
0
1
0 1
"Gnathotrusia mundina
3
?
1
?
0
0
0
0
0
0
?
0
1
0
0
0
0
1
0 1


415
Strict
d
CI=. 779
RI=. 913
RC=.712
d
d
C. nycteis
C. gorgone
C. n. sp. nr. har.
C. harrissii
C. hoffmanni
C. palla
C. gabbii
C. acastus
C. w. damoetas
C. w. whitneyi
C. deflnlta
T. ezra
T. theona
T. perlula
T. chinatiensis
T. I. leanira
T. I. fulvia
T. cyneas
T. cynisca
C. e. pardelina
C. e. endels
C. melitaeoides
C. marina
C. e. poecile
C. e. erodyle
C. melanarge
C. eumeda
C. californica
C. lacinia
C. ehrenbergi
C. hippodrome
C. narva
C. g. gaudealis
C. g. wellingi
C. janais
C. marianna
C. rosita rosita
C. montana
C. rosita browni
C. riobalensis
C. mazarum
T. eleda
T. coracara
T. anomalus
D. dymas
M. elva
A. pelops
A. proclea
P. minuta
H. fasciatus
H. miriam
A. perezi
A. pantoni
A. tulita
CumPhyGnaMel
Figure 320: Strict consensus tree from a heuristic search of successively weighted characters from the
Chlosyniti/Poladryiti data matrix with the DPCWH method for characters with multistate taxa.


580
presence of an orange patch at the inner anterior margin of the eye could not be
determined due to damage to the specimens in this area. The femur of the meso- and
metathoracic legs is orange dorsally and white ventrally as in C. chinatiensis, however in
one of the two specimens that had one of these legs in reasonably good condition a row
of gray scales bordered the white scales. The pattern of the mesopleurons where the
femur tucks in could not be determined from any of the material examined. Two
specimens had intact antennae which lacked orange scales, but other taxa in the
monophyletic group of taxa formerly classified as Thesscilia have orange scales present in
a portion of individuals, which may well be the case with C. perlula as well.
Wing fringe checkered black and white with the white sections present between
the veins, and the white section adjacent to forewing cell Ml relatively small. Forewing
basal costal area with a thin row of orange scales.
Dorsal forewing with the same symmetry system bands present in the same cells
as for C. chinatiensis (above) but in all examples examined these markings are very
diffuse. The basal band of the basal symmetry system in the discal cell is strongly curved
(convex in distal to basal aspect) forming a basal patch that may or may not have diffuse
orange scaling inside. The patch at the distal end of the discal cell is a contrastingly light
cream color against the orange background, and does not extend all of the way to the
posterior border of the discal cell. The median band patches are large and a contrasting
cream color, much more so than in C. chinatiensis. Distal to the median band patches in
cells CuA2-CuAl, the wing is predominately black, and the black area also occurs basal
to the median band in cells M3-R3. The black area extends to the posterior distal edge of
the discal cell, but the basal parts of cells M3 and M2 still contain patches of orange. The


704
Species Delimitation: Chlosyne gaudealis has consistent discontinuities in male genitalic
and wing pattern variation with its sister taxon, C. narva, and the ranges of these taxa
overlap.
Subspecies Delimitation: I recognize two subspecies of C. gaudealis, C. gaudealis
gaudealis and C. gaudealis wellingi. I have seen no intermediates between the nominate
gaudealis phenotype and the wellingi phenotype, and according to Miller and Rotger
(1979), the two phenotypes are allopatric. Phylogenetic evidence indicates wellingi and
gaudealis are sister taxa, and the only character difference between them is the size of the
forewing red patch.
Chlosyne gaudealis gaudealis
Diagnosis: Nominate gaudealis is separated from subspecies wellingi by the presence of
red to red-orange scaling on the dorsal forewing surface.
Further Description: The entire dorsal forewing is fdled with red to dark red-orange
scaling. The extent of the red scaling in adjacent areas varies along a continuum, and
may include a patch in cell CuA2, a small patch in the basal most part of cell CuAl, and
patches in the basal parts of the cells adjacent to the distal end of the discal cell. A
minute disjunct red patch may occur in cell R5. The red patches are separated only by
thin amounts of black scaling along the veins. On the ventral forewing surface, the red
patches occur in the same position as on the forewing or are more reduced (patches
outside the discal cell present on the dorsal surface may be reduced or absent on the
ventral surface) in contrast to subspecies wellingi.


250
band (see below); I coded this taxon state 0&1 for the following character. Likewise,
only one specimen of a larger series of C. ehrenbergi had a few orange scales on its labial
palpi.
53. Presence of orange hairs and scales on the labial palpi.
0=Present.
l=Absent.
@A=0&1: Present in some individuals but absent in others.
54. Presence of black hairs and scales on the labial palpi.
0=Present.
#l=Absent.
55. Dorsal surface of labial palpi.
0=Predominantly black scales/hairs with scattered orange and white hairs and
scales.
l=Predominantly black scales/hairs with scattered orange but no white scales and
hairs.
2=Predominantly orange scales/hairs with scattered black hairs.
3=Covered with black scales and hairs only.
4=Covered with orange scales and hairs only (except some sparse grey-black
hairs on proximal most end).
5=Predominantly orange scales/hairs but with some black hairs near base.
#6=With orange scales and hairs except for a band of black scales and hairs
running along the dorsal midline and some black scales at the apex. A variable
amount of scattered black scales may occur to the sides of the black band.


173
3 OX
23(0)
40X
23(0)
155
Figures 154-155. Male genitalia capsule of Chlosyne sp. In ventral aspect. 154: C.janais.
155: C. rosita. Figs. 156-157. Male genitalia capsule of Chlosyne sp. In dorsal aspect.
156: C.janais. 157: C. rosita.
156
154
Ventral
valve
3 OX
Posterior valve


245
Taxa lacking state 0 are coded *=? The presence of inverted teeth on the corpus bursae is
a symplesiomorphy.
42. For those taxa with character 16 state 0, the sclerotization of the teeth on the corpus
bursae:
0=Appearing non-sclerotized, concolorous with the corpus bursae.
1 ^Distinctly sclerotized, and often with a variable amount of sclerotized tissue on
the corpus bursae between the teeth (Figures 284,286 & 289-291).
Taxa lacking inverted teeth on the corpus bursae are coded *=?.
All Melitaeini have a sclerotized plate on the ventral surface of the corpus bursae:
43. Paired anterior extensions on the ventral corpus bursae plate:
#0=Absent (Figures 274 & 276).
l=Present (Figures 278-284 & 286-292).
44. Presence of inverted teeth on the ventral corpus bursae plate.
#0=Absent.
#l=Present.
45. Given state 0 of the second preceding character, the shape of the ventral corpus
bursae plate:
l=Quadrate, at most 3 times as long as the width of the anterior side, posterior
side widest ad slightly wider than the anterior side (Figure 276).
2=About as wide on the anterior side as long, anterior side about 2X the width of
the posterior side, sides widen in a concave curve anteriorly with a convex bulge
nearest the anterior side, a short triangular extension of the plate extends
anteriorly along the ventral midline (Figure 274).


716
from Jalisco (Tenacatita), a state from which I have seen no specimens of C.janais. I
have seen the marianna phenotype from the Mexican states of Guerrero, Morelos,
Puebla, and Veracruz. It appears to be locally common in some areas, and the AM series
include over 135 specimens from Guerrero (Mexeala, Acahuizotla, and Acapulco) and
Morelos (Rancho Viejo, R. de la Maza, Paraje Rancho Viejo). There is also a single
specimen labeled from the Canal Zone in Panama, which perhaps should be viewed with
suspicion absent further corroboration. I have seen specimens representing intermediate
phenotypes between the nominate janais, gloriosa, and marianna phenotypes from the
Mexican states of Guerrero (Acahuizotla and Mexeala) and Michoacan (Coahuyana).
Higgins (1960) reported Jalisco (Tenacatita) as another locality where intermediate
phenotypes have been collected. The nominate janais phenotype occurs along with the
other phenotypes in Guerrero (Acahuizotla) (examined) and apparently also in Jalisco
(Tenacatita) (Higgins 1960).
Species Delimitation: Chlosyne janais has discontinuities in male genitalia and wing
pattern, compared with all other Chlosyne. Phylogenetic evidence indicates C.janais is
most closely related to C. rosita, which occurs within the range of C. janais without any
evidence of integredation, although I am uncertain if the two taxa ever fly together in the
same locality at the same time. There is no evidence to suggest the marianna and
gloriosa phenotypes are anything more than part of the continuum of variation which
exists for the C. janais lineage, as they occur together with the nominate janais
phenotype with every intermediate phenotype exhibited as well.
Chlosyne rosita Hall


533
diffuse scaling. Males vary along a continuum from being very orange predominately
black in the basal areas, but the postmedian bands remain orange in all specimens.
Wing Span: S'- 30-39mm. $:39-46mm.
Geographic Variation: C. palla varies along a continuum with respect to the extent of
black and color of light markings on the dorsal and ventral wing surfaces. The series in
the National Museum of Natural History show a continuum of integredation between
lighter and darker specimens and between deep orange and pale cream light markings,
respectively, in both females and males. The same is true of the ventral hindwing
background color, which varies along a continuum between yellowish cream and pale
cream In series from some areas there are a mix of light and dark females, while in other
series females are all light or all dark. In the long series of C. palla in the NMNH there is
an impressive amount of variation, but I fail to failed to detect a gap in the variation of
any character; hence, I recognize no subspecies. Without doubt, different populations
have differences is the average variation of some wing pattern characters and the range of
variation expressed for those characters (particularly the amount of black on the dorsal
wing surfaces, sexual dimorphism, color of the light markings on the dorsal wing
surfaces, contrast between the postmedian band and median band, and shade of cream
coloration of the ventral hindwing background color), and the geographic variation is
complex rather than falling along a simple cline.
J. Emmel et al.(1998d) recognize the greatest number of C. palla subspecies,
including palla Boisduval,y7avi//c/ Barnes and McDunnough, calydon Strecker, sterope
Edwards, eremita Wright, altasierra Emmel, Emmel, and Mattoon, and australomontana
Emmel, Emmel, and Mattoon. These authors report being aware of blends or


7
showing that the initial hypothesis is no longer favored, as opposed to arbitrarily deciding
that Taxon A is more similar to Taxon C. Even in a natural classification, disputes can
still arise over the validity of particular taxa as two workers may reach conflicting
hypotheses from the same evidence. However, the scientific method is self-correcting
over time, as new evidence and/or an improved theoretical framework favors one
hypothesis over its alternatives.
In all aspects of this work, I choose the criterion of monophyly as a universal
requirement for the validity of a taxon of any rank. While monophyly provides a
scientific basis for testing the validity of taxonomic concepts, it does not address another
issue in the systematic classification of organisms: how many natural taxa does one
formally name and how are they ranked. Above the species level, I see no way such a
question can be answered with the scientific method, and the issue is a matter of
bookkeeping rather than science. However, I argue that making such bookkeeping
decisions objectively is preferable to arbitrarily doing so. This goal can be achieved by
applying a secondary criterion of nomenclatural stability, as discussed in Chapters 2 and
3 where phylogenetic evidence is used to convert the existing classification schemes into
a natural classification scheme.
Species and Subspecies Concepts Used in this Work: Theoretical Concepts of the
Species/subspecies Category and Practical Applications for Delimitating
Species/subspecies Taxa
The Species Problem:
Ereshefsky (1992) makes an important distinction regarding the use of the term
"species." One usage is the species category, relating to what characteristics make an
entity a species, and the other usage is species taxa, that which is formally named as a


599
Figures 146-147,167,169,185,199,225,248 & 284.
Melitaea leanira C. & R. Felder, 1860. Wiener Ent. Monats. 4:106. Type Locality:
"California", restricted to Hwy 70 at Chambers Creek, North Fork Feather River
Canyon, 1850', Plumas County, California by J. Emmel et al. (1988b). Lectotype:
British Museum (figures 13-15 in J. Emmel et al. (1998b)).
Diagnosis: I recognize four distinct subspecies within the specific taxon Chlosyne
leanira: C. leanira leanira (C. & R. Felder), C. leanira fulvia W. H. Edwards, C. leanira
cyneas (Godman and Salvin), and C. leanira cynisca (Godman and Salvin). Like other
taxa formerly classified in the genus Thessalia, the ventral hindwing has thick black
scaling along the veins, the hindwing basal symmetry system is absent, both the basal and
distal median lines are absent, and the posterior valve projection points inward in
posterior view. All of the C. leanira subspecies may be easily separated from all taxa in
the Chlosyne ezra clade by the cream colored dorsal hindwing postmedian band
(sometimes appearing as postmedian dots) and the absence of orange in the ventral
hindwing postmedian band. Also, the ventral hindwing discal cell contains a black
parallel line in part of its center which forks distally and fuses with the black border along
each wall of the discal cell, but this line is not complete in all specimens.
The genitalic differences between C. leanira and the C. ezra clade are discussed
under the diagnosis of C. ezra. The four subspecies of C. leanira can not be
differentiated by genitalic characters. The orientation of the inner valve process is
variable within C. leanira, and the process may be orientated such that the posterior side


327
Some taxa have a variable (geographically or not) amount of solid black
extending into the basal part of the wing, such that some specimens cannot be scored for
the band (which is all black in these taxa) but others can. For example, some specimens
of Chlosyne rosita with the montana phenotype are heavily marked with black in the
basal area, but other C. rosita specimens clearly show the band is in the form of state 5.
Coding for these taxa is based on only those specimens with enough light background in
the basal area to score this character. Taxa where all individuals are solid black in this
area are coded
137. Some taxa have the above ventral hindwing band variably broken up into a series
of disconnected patches (how much of the band is broken up and the size of the breaks is
intraspecifically variable). This is to be distinguished from the less coalesced patches of
Microtia and Polyclryas, since the condition it is due to the loss of sections of the band
rather than the sections being spread farther apart.
For those taxa with a band formed between the proximal and distal elements of the
central symmetry system, the completeness of band sections:
0=Cells contain complete sections, forming a continuous band unless the sections
are poorly coalesced.
l=Cells containing incomplete sections, band comprised of a series of disjunct
patches.
Taxa lacking the above band, or with the background to dark to detect it, are coded
138. For those taxa where some form of the ventral hindwing band in the preceding
character is present, the color of scales comprising the proximal and distal elements of
the central symmetry system:


37
2=A process of variable shape thats distal most point curves anteriorly
(sometimes only slightly) with the anterior wall of the distal most point concave
and the posterior wall convex (Fig. 36).
Taxa lacking a ventral valve process are coded 0.
16. For those taxa with a ventral valve projection of character 14, the variation in shape
of the process among state 1 and state 2 taxa for the preceding character can further be
divided into discrete states:
l=Curved and pointed with the walls entire (or with one tooth-see the following
character) (Figs. 1,85,91,96,101,114,116,118 & 120).
2=Anterior edge smooth with 1-2 small teeth distally, posterior edge serrate with
several (approximately 6) distinct teeth (Fig. 82).
3=Very jagged, anterior portion of process with 2-3 prominent teeth in a row
(some additional smaller teeth may be present in-between) with a similar distal
extent, posterior portion of process extends much farther distally as a curved,
pointed projection with 0-1 smaller teeth on its anterior edge and 1-2 smaller teeth
on its posterior edge (posterior to the longest prong there may be 3 or 0 short
teeth) (Fig. 36).
4=A deeply forked projection, with the longest (anterior) fork curving anteriorly
and the shorter (posterior) fork curving posteriorly.
5=A single broad process with a serrate anterior side and an entire posterior side.
States 1 and 2 are for taxa which have state 1 for the preceding character, while states 3-6
are for taxa which have state 2 for the preceding character. The above coding avoids
biasing the analysis for or against the homology between these two states of character 14,


283
3=Deep yellow.
*#4=A mosaic of orange and cream colored scales with sparse black scales
interspersed.
@A=0&2: Varies along a continuum between state 0 and state 2.
@*B=1&2: Some individuals are state 1 while others are state 2.
@*C=2&3: Varies between state 2 and state 3.
Intraspecific variation of this patch in females of C. palla and C. hoffmani varies along a
continuum from orange to cream colored to almost white (but a distinct gap occurs
between this off white and the pure white of state 1), while all the other taxa can be coded
as either orange or cream colored. Since a distinct gap exists between orange and cream
for all other taxa, and C. palla and C. hoffmani exhibit both states with intermediates, I
elect to code C. palla and C. hoffmani as 0,2.
I code this character separately for the two sexes since there are a number of
sexually dimorphic taxa, and the sexual dimorphisms among these taxa are not all
equivalent. The variation between the two characters appears to be primarily
independent and non-reinforcing.
92. For those taxa with the patch of the third preceding character, the separation or
continuity of this patch on the dorsal wing surface:
0=A single patch of light colored scales is present.
l=The patch is distinctly separated into an upper and lower element.
@A=0&1: Varies from a continuous patch (tending to be smaller than that of
most state 0 individuals) to two disjunct patches (tending to be larger than most
state 1 taxa).


Table 12 Continued
740
NMNH=Nat Mus of Natural Hist -Smithsonian Insl FSCA=FL St Colleclion ol Arthropods AM-Allyn Mus PC=Personal Collection @=Judqed by author as ol dubious authenticity
County (USA)
Locality (USA)
Taxon
Country
State/Prov.
Locality (non US)
Locality Continued (non USA)
Date
Collector
Collector
No
Chlosyne theona
Mexico Guerrero iguala
June T)6
NMNH
,
Chiosyne theona
Mexico Hidalgo 60 m. N of Pachuca, 8500 ft
1 Sept 1971
Robbins
NMNH
3
Chlosyne theona
Mexico Hidalgo Sabinas
7 III 36
C L Gibson
NMNH
1
Chlosyne theona
Mexico Hidalgo Sabinas
24 III 36
C L Gibson
NMNH
4
Chlosyne theona
Mexico Hidalgo 109 mi Sol Tamazunchale
NMNH
1
Chlosyne theona
Mexico Morelos Cuernavaca
W Schaus
NMNH
1
Chlosyne theona
Mexico Morelos Cuernevaca
Aug. '06
W Schaus
NMNH
1
Chlosyne theona
Mexico Nuevo Leon Horsetail Falls
19 Jun 1986
I L Finkelstein
FSCA
t
Chlosyne theona
Mexico Nuevo Leon Hwy 85 c. 40 mi S Nuevo Laredo
26 Jul 1976
I L Finkelstein
FSCA
1
Chlosyne theona
Mexico Nuevo Leon Monterrey
Jun.5/35
NMNH
1
Chlosyne theona
Mexico Nuevo Leon Monterrey
VI-5-35
NMNH
1
Chlosyne theona
Mexico Nuevo Leon Monterrey
28 May 1941
NMNH
1
Chlosyne theona
Mexico Nuevo Leon 100 mi N of Monterrey
NMNH
1
Chlosyne theona
Mexico Oaxaca Oaxaca
W M Schaus
NMNH 1
Chlosyne theona
Mexico Oaxaca Monte Alban
3 Aug 1976
I L Finkelstein
FSCA
Chlosyne theona
Mexico Oaxaca 43 Mi W of Tehuantepec. 2500-3000 ft
27-VIII-63
FA Lee
FSCA
2
Chlosyne theona
Mexico
Oaxaca 18 km W of Tehuar
San Luis Potosi ei Salto
8/19/80
EC Olson
FSCA
!
Chlosyne theona
Mexico
3-VII-1968
Debra Weems
FSCA
1
Chlosyne theona
Mexico San Luis Potos ei sato Fails. 2000-2500 ft
8 VIII 63
H V Weems
FSCA
Chlosyne theona
Mexico San Luis Potosi ei Salto Fails
20-X-1970
H L King
FSCA
,
Chlosyne theona
Mexico San Luis Potosi Hwy 85
30 July 1976
ILF
FSCA
1
Chlosyne theona
Mexico San Luis Potosi Paiiua
4-15-1972
H L King
FSCA
.
Chlosyne theona
Mexico San LUS PotOSi Tamazunchale
10-15 Apr 1949
EL Todd
NMNH
:
Chlosyne theona
Mexico San LUS PotOSi Tamazunchale
12 IV 1949
EL Todd
NMNH
5
Chlosyne theona
Mexico San Luis Potosi Tamazunchale
30 July 1976
I L Finkelstein
FSCA
2
Chlosyne theona
Mexico San Luis Potosi 10 mi SW Tanazunchale
12 Aug 1972
GF&SHevel
NMNH
Chlosyne theona
Mex ico San Luis Potosi 13.8 mi S of Tamazunchale
21 VIII 1967
Gary F Hevel
NMNH
Chlosyne theona
Mexico San Luis Potosi 16 Mi S of Tamazunchale
18 Aug 1967
Gary F Hevel
NMNH
Chlosyne theona
Mexico San Luis Potosi vaiies
NMNH
1
Chlosyne theona
Mexico Sinaloa
NMNH
)
Chlosyne theona
Mexico Sinaloa venad.0
W M Schaus
NMNH
7
Chlosyne theona
Mexico Sonora Mazatlan
20 July 1956
Mai Douglas
FSCA
1
Chlosyne theona
Mexico Tamaulipas Ciudad Mante
H L King
FSCA
1
Chlosyne theona
Mexico Tamaulipas
33 mi N of Cd. De Maite
17-VIII 1967
Gary F Hevel
NMNH
1
Chlosyne theona
Mexico Tamaulipas
2 Mi SW of Ciudad Mante. 0-500 ft
7-VIII-63
HV Weems Jr
FSCA
Chlosyne theona
Mexico Tamaulipas
Cd. Valles
19-X-1970
H L King
FSCA
2
Chlosyne theona
Mexico Tamaulipas
Galaena Canyon
16-VIII 1967
H L King
FSCA
Chlosyne theona
Mexico Tamaulipas
Nr Gomez Farias
March 5-8 1959
T C Emmel
FSCA
,
Chlosyne theona
Mexico Tamaulipas
San Fernando
15-V-1973
H L King
FSCA
3
Chlosyne theona
Mexico Veracruz
Cordoba
VIII 07
R Muller
NMNH
,
Chlosyne theona
Mexico Veracruz El Tarjin Ruins. 10.5 Mi S of Poza Rica
11-VIII.63
HV Weems Jr
FSCA
,
Chlosyne theona
Mexico Veracruz Fortin de las Flores. Sumidero Hydroelectric Plant

Daniel Rabago
FSCA
1
Chlosyne theona
Mexico Veracruz Jalapa
NMNH
2
Chlosyne theona
Mexico Veracruz j^apa
B Neumoegen
NMNH
,
Chlosyne theona
Mexico Veracruz Jalapa
W Schaus
NMNH
,
Chlosyne theona
Mexico Veracruz LagoCatemaco
11-X-1970
H L King
FSCA
2
Chlosyne theona
Mexico Veracruz oizaba
26 Aug 1969
William H Howe
NMNH
,
Chlosyne theona
Mexico Veracruz Pueblo Viejo
XII.8.09
NMNH
2
Chlosyne theona
Mexico Veracruz Rmco-NadaSE of Jalapa
13-VIII 63
FA Lee
FSCA
,
Chlosyne theona
Mexico Veracruz Santa Rosa
May 06
W M Schaus
NMNH
Chlosyne theona
Mexico Veracruz Veracruz
VIII-59
JCHopfmger
NMNH
2
Chlosyne theona
Mexico Yucatan
Oberthur
NMNH
2
Chlosyne theona
Mexico Yucatan Piste
1 Sept. 1967
NMNH
,
Chlosyne theona
Mexico Yucatan Piste
29 Aug 1968
NMNH
,
Chlosyne theona
Mexico Yucatan Piste
2 Sept 1968
NMNH
,
Chlosyne theona
MexiCO ? RioS Diego
VII.07
NMNH
,
Chlosyne theona
Mexico ? Sierra Blanca (sp?)

J C Hopfmge.
NMNH
,
Chlosyne theona
Mexico ? Sierra Planea
VIII 41
J C Hopfinger
NMNH
2
Chlosyne theona
MexiCO ? Thomuzunchah [misspelled Tamazunchale, San Luis Potosi?)
July-1957
GW Rawson
NMNH
2
Chlosyne theona
MexiCO ? Thomuzunchali
4 July 1958
G W Rawson
NMNH
l
Chlosyne theona
Mexico ? Mexique (French for Mexico)
NMNH
Chlosyne theona
Mexico ? Mexique (French for Mexico)
Oberthur
NMNH
3
Chlosyne theona
Guatemala Baja Verapaz San Geranano de Jutiama
24-11-71
H L King
FSCA
Chlosyne theona
Guatemala El Peten Tikai ruins
2 Sept 1972
G F & S Hevel
NMNH
1
Chlosyne theona
Guatemala Zacapa Zac¡*>a
NMNH
Chlosyne theona
Guatemala ? Laquin. 300m
3 Nov. 1985
EC Olson
FSCA
Chlosyne theona
Honduras Atlantida 18 km w of LaCetba
22-IV-1981
Robert Lehman
NMNH
,
Chlosyne theona
Honduras Copan copan Rums
17 Aug 1974
Robert D Lehman
NMNH
,.T
Chlosyne theona
Honduras Cortes Sm Pedro Sula. roadside
8-IV-1972
Robert Lehman
NMNH
Chlosyne theona
Honduras ? Augustine
XI-27-67
FSCA
Chlosyne theona
Honduras ? Orange Walk
30-X-1967
H L King
FSCA
Chlosyne theona
Honduras ?
EdwTOwen NMNH
Chlosyne theona
Nicaragua Chinandega San Marcos
Chlosyne theona
Nicaragua Managua Managua
20 July 1974
R Anderson
NMNH
,
Chlosyne theona
Nicaragua Managua Managua
21 July 1974
R Anderson
NMNH
Chlosyne theona
Nicaragua Managua Managua
28 Aug 1975
R Anderson
NMNH
Chlosyne theona
Nicaragua Managua Managua
19 Oct 1975
R Anderson
NMNH
2
Chlosyne theona
Nicaragua Managua Managua
15 Nov. 1975
R Anderson
NMNH
,
Chlosyne theona
Nicaragua Managua Managua
27 Nov 1975
NMNH
Chlosyne theona
Nicaragua Managua Managua
1 Aug 1976
R Anderson
NMNH
Chlosyne theona
Nicaragua Managua Managua
5 Sept 1976
R Anderson
NMNH
Chlosyne theona
Nicaragua Managua 7 m se Managua
20 Dec 1974
R A Anderson
NMNH
Chlosyne theona
Nicaragua Managua 15 km s Managua
22 Oct 1976
NMNH
1
Chlosyne theona
Nicaragua ? 31 m n of Esteii
22 Aug 1972
G F & S Hevel
NMNH
Chlosyne theona
Nicaragua ? Granada
23 Aug 1975
Chlosyne theona
Nicaragua ? 5 Mi SE Granada
31 July 1976
R A Anderson
NMNH
Chlosyne theona
Costa Rica Cartago Tres Rk>s. sooort
Dec-06
NMNH
Chlosyne theona
Costa Rica Guanacaste 5km NW Caas Hac La Pacifica. 50m
VIII-26-73
Chlosyne theona
Costa Rica Guanacaste FmcaComeico
16 Jan 1970
Chlosyne theona
Costa Rica
Guanacaste Finca Comelco
26-IX-1970
2
Chlosyne theona
Costa Rica
Guanacaste La cruz
26-IX-1970
H L King j FSCA
Chlosyne theona
Costa Rica
G uanacaste Fincha Pacifica Canas
1 July 1970
M L May FSCA
Chlosyne theona
Costa Rica
Heredia San Antonio de Belen, 900m
7 VIII 1977
G B Small NMNH
Chlosyne theona
Costa Rica
San Jos Purisca Mts
W M Schaus NMNH
Chlosyne theona
Costa Rica
San Jos
EdwTOwen NMNH
,
Chlosyne theona
Costa Rica
San Jos Uooo'
Nov '06
Chlosyne theona
Not Located
2 km E of Raya Azul
5/3/75
Chlosyne theona
Not Located
Tecalpa
J-7-67
Chlosyne theona
? [Prob USA]
7 V 1967
j
Chlosyne theona
7 tpfb USA1
SWRS
FSCA


712
median band patches dorsally, but in specimens where the median band is incomplete
(not including cell M3) dorsally additional patches are often present ventrally. The
gloriosa phenotype often has the dorsal median band confined to cells M2-R3 where the
patches are enlarged, but small patches are usually present in some or all of the remaining
cells (except for M3) ventrally. The light patch in the discal cell basal to the position of
the discal spot occurs as two disjunct white dots, which are absent dorsally (but present
ventrally) in some specimens of the gloriosa phenotype and almost always present
dorsally as well as ventrally in the other phenotypes. An additional small white patch is
almost always present dorsally and ventrally in the discal cell, located along the anterior
vein basal to the position distal band of the basal symmetry system. A small white patch
is often present dorsally in cell CuA2 just basal to the origin of vein CuA2, and when
absent on the dorsal surface it is often present ventrally. Some specimens have very
diffuse scaling comprising weak pale to yellowish cream colored patches distal to the
postmedian dots on the dorsal surface, however these patches are sharp and conspicuous
ventrally in at least some cells, and usually in cells M3-R5 with smaller patches
sometimes present in cells CuA2-CuAl and R4. The M3 patch is always the largest
followed by the M2 and R5 patches, and these three patches are apparently always
present ventrally (although they may be very small in some specimens). A narrow area
of diffuse cream scaling is usually visible dorsally in the extreme basal part of the discal
cell (at least with magnification), and sharper and larger in the same position ventrally.
The ventral forewing is like the dorsal surface except the light colored markings are
larger, and usually more numerous in specimens where the light markings were
incomplete on the dorsal surface.


205
Bootstrap
87
71
81
94
100
76
81
81
77
100
81
88
69
99
88
57
63
83
87
57
72
75
84
99
66
63
85
74
91
85
83
100
86
80
83
59
88
77
81
83
97
75
Chlosyne A
Chlosyne B
T. elada
M. elva
D. dyrnas
T. corocara
T. anomalus
A. pelops
A. proclea
P. minuta
H. fasciatus
H. miriam
A. tulita
A. pantoni
M. athalia
M. britomartis
M. alatauica
M. parthenoides
M. varia
M. aurelia
M. asteria
C. phoebe
C. scotosia
M. cinxia
M. diamina
D. didyma
D. trivia
P. tharos
P. batesi
P. phaon
P. simois
A. texana
A. tulcis
T. teletusa
C. castilla
D. catula
O. liriope
O. ithra
E. eunice
J. leucodesma
J. fellula
M. amaznica
T. claudina
T. saladellensis
G. exclamationis'
G. mundina
G. sodialis
E. aurinia
E. desfontaini
H. matuma
H. intermedia
H. gillettii
H. cynthia
H. iduna
O. anicia
O. chalcedonia
O. colon
E. phaeton
O. editha
A. amathea
A. fatima
A. jatrophe
C. cloanthe
J. coenia
Chlosyniti
Poladryiti
Melitaeiti
Phycioditi
Gnathotrichiti
Euphydryiti
Figure 295: Boot strap 50% consensus tree for equally weighted characters from the Melitaeini
data matrix with the representative out group method.


472
distal median line and the distal band of the central symmetry system (contrast Antillea).
When present and detectable against the background, the section of the distal band of the
central symmetry system in cell CuAl is not aligned with the sections in cells CuA2 and
M3.
Remarks: In addition to the generic synonyms listed above, taxa which I place within the
Microtia have historically also been classified in the following genera: Melitciea
Fabricius, Phyciodes Hubner, and Eresia Boisduval. These genera are not synonyms of
Microtia because their type species are members of a different clade.
Microtia eleda (Hewitson)
Figures 108-109,114-115,128-129 & 280.
Eresia eleda Hewitson, 1868. Exot. Butt. 4: Eresia, pi. VII, figs. 54,55. Type Locality:
Mexico. Holotype: British Museum (Higgins 1960).
=Melitaea callina Boisduval, 1869. Ann. Soc. ent. Belg. 12:54. Type locality: Sonora,
Mexico. Syntype: Carnegie Museum, Pittsburgh (Figures 231-233 in J. Emmel et
al. 1998d), Syntype: NMNH [Examined],
=Eresia soda C. and R. Felder, 1869. Verh. zool.-bot. Ges. Wien. 19:470.
=Melitaea Ulrica W. H. Edwards, 1877. Caad. Ent. 9:189. Type Locality: San Antonio,
Texas. Lectotype: Carnegie Museum (Brown 1966).
=Melitaea immitata Strecker, 1877. Lepidoptera Part 14:130. Type Locality: Texas.
Holotype: Chicago Nat. Hist. Museum (Higgins 1960).


652
Range: C. erodyle erodyle ranges from Mexico (Jalapa and Mexico City and south) to El
Salvador and Honduras, and there is one specimen in the NMNH labeled "Texas". The
distribution of specimens examined with more data than the country includes the Federal
District (Mexico City), Chiapas (San Jeranimo, Rizo de Oro, and 25 miles east of Tuxtla
Gutierrez),Veracruz (Jalapa, Orizaba, and Paso San Juan), Yucatan (Piste and Ruinas
Ozibilchaltun), Honduras (Cayo District: Camp Sibun), Guatemala (Retalhulea,
Escuintla, Lake Amatitlan, and Olas de Moka Dept. Solola), El Salvador (San Salvador,
2km N of San Isidro). Localities not located include Sierra Blanca and Mirador; cities
which I found with these names are not within the expected distributional area. Once
again there is an odd record from the Barnes collection, for Chiriqui, Panama, which I
discard absent corroboration. Specimens for which data was recorded are from the
NMNH and FSCA, and detailed collection data appears in Table 12. Higgins (1960)
reports records for Nicaragua (Chontales and San Ramon).
Chlosyne erodyle poecile (C. and R. Felder)
Figure 143.
Synchloepoecile C. and R. Felder, 1867. Reise Novara. 3:396. Type Locality: Bogota-
New Granada. Holotype: British Museum.
Diagnosis: See the diagnosis of Chlosyne erodyle erodyle above for the only characters
which differ between subspecies poecile and subspecies erodyle.


557
Melitaea definita Aaron, 1884. Papilio. 4:176. Type Locality: Inland from Corpus
Christi, Texas. Syntypes: Chicago Field Museum of Natural History (Miller and
Brown 1981).
-Melitaea albiplaga Aaron, 1884. Papilio. 4:175. Type Locality: Inland from Corpus
Christi, Texas.
=Melitaea schausi Godman and Salvin, 1901. B. C. A. 2:676. Type Locality: Vera
Cruz, Mexico. Holotype: British Museum (Higgins 1960).
-Melitaea beckeri Godman and Salvin, 1901. B. C. A. 2:676. Type Locality: Durango,
Mexico. Holotype: British Museum (Higgins 1960).
=Melitaea anastasia Hemming, 1934. Stylops. 3:193. Prpoposed as a replacement for
beckeri Godman and Salvin (Miller and Brown 1981).
Diagnosis: Chlosyne definita is most likely to be confused with C. encleis and to a lesser
extent C. marina, taxa which may have a similar form of the ventral hindwing
postmedian band and of other characteristics on the dorsal and ventral wing surfaces.
Both C. definita and C. endeis are variable with respect to the dorsal wing pattern. In C.
definita orange scales are present on the vertex and dorsal surface of the abdomen, while
in C. endeis they are not. Dorsally labial palpi predominately orange (with scattered
balck) in C. definita but just black in C. endeis. Sutures at the lateral edges of the vertex
with orange, white, and black scales in C. definita but just black ones in C. endeis', the
same is true of the collar between the head and pronotum except C. endeis also has white
scales. The bands of the symmetry systems are distinct black against an orange
background ventrally in C. definita, while they are absent in C. endeis where the ventral


228
#0=With an elongate tapering extension with the sides slightly concave (Figures
88,92 & 97-98).
l=With a broad extension with a convex base (Figures 127,129,131,133, & 213-
233).
Within state 1 taxa there is variation as to whether the sides are convex all the way to the
tip, or whether they straighten out more distally. I found this variation could not be
coded into discrete states, since there were no obvious gaps and some intrataxon
variability occurs, although there may be differences in the average extent or range of
variation between some state 1 taxa.
5. For those taxa with state 1 of the preceding character, the posterior end of the phallus
extension:
l=Broadly squared off (Figures 229,232 & 233).
2=Curving to a (usually blunt) point (Figures 104,127,129,131,133,213-228 &
230-231).
Independent evidence from other characters suggests state 1 of the preceding character is
a terminal derived state. In any event, the out group taxa with state 0 of the preceding
character have the distal end of the phallus extension narrowly rounded off. These taxa
are coded 0.
6. For those taxa with character 4 state 1, the presence of a centered terminal triangular
extension with concave sides attached to the distal end of the broad convex sided
extension:
l=Present (Figure 104).
2=Absent.


560
Narrow band of orange scales along the base of the forewing costal margin. Wing
fringes checkered black and white with white sections between where the veins reach the
wing margin. Basal part of ventral forewing cell C orange and continuous with
surrounding orange and sometimes with white in the upper part of the cell except at the
extreme basal area. Discal spot present with orange scaling inside. Forewing discal cell
with a patch fonned from the fusion of the distal band of the basal symmetry system and
proximal band of the central symmetry system with orange scaling inside the same color
orange as that inside the discal cell. Dorsal forewing discal cell with a continuous or
divided cream colored patch between the proximal band of the central symmetry system
and the discal spot, with a variable number of orange scales around the periphery (the
peripheral orange scales tend to decrease southward, and are absent in some Mexican
specimens). Basal band of basal symmetry system forming a basal patch in the discal cell
with diffuse orange scaling of identical color to the orange inside the discal spot. Dorsal
forewing cell CuA2 with the distal element of the basal symmetry system single and
forming a narrow, elongate patch with diffuse orange scaling inside extending from the
basal part of the cell almost to the origin of vein CuA2. Dorsal hindwing postmedian
band present and in the form of a series of light patches in cells CuA2 to R5 or to Sc+Rl,
patches either orange in each cell or with cream patches in cells M3, R5, and Sc+Rl (if
present). Ventrally patches in cells CuA2-Sc+Rl with orange patches in CuA2-CuAl
and M2-M1, a cream patch in M3 with or without an orange periphery, and cream
patches in R5 and Sc+Rl (the cream patches are the same color as the cream background
color (not including the postmedian dot in cell M3 which may occupy most of the patch
and be a slightly different shade of cream). Forewing postmedian band of orange patches


21
out group taxon exhibits a state that occurs in the in group: the out group is coded c
(the number of states occurring in out group taxa, or how they would be designated is
irrelevant to cumulative out group coding). 4) Some out group taxa exhibit a state that
does not occur in the in group, while others exhibit one or more states that do occur in the
in group (n and p): the out group is coded c,n,p. (5). Some out group taxa are coded
for a particular character, but state information is unknown for one or more taxa included
in the cumulative out group: the out group is coded 6) A character clearly falls into
discrete states (n and p) within the in group, but not among out group taxa: the out group
is coded ?, while in group taxa are individually coded n or p as appropriate.
In an attempt to include a sample representative of the character state variation
within Nymphalini and Kallimini, I examined the males of the type species of each
Nymphalini and Kallimini genus listed in Harveys (1991) Table B.2 except for a few
genera for which I did not have males of the type species available (Bassctris,
Symbrenihia, and Mynes). Information on the specific types for these genera was
obtained from Hemming (1967). I also included a representative of Liminetidini,
Colobura clirce, in the out group. The taxa comprising the cumulative out group are
listed in Table 1 under Out Group Taxa. If the cumulative out group method is applied
to the phylogenetic analysis of a genus where several genera are potentially sister taxa,
for scoring the cumulative out group it would be best to include all species taxa in the
genera which are potential sister taxa. However, with respect to the phylogenetic
analysis of a tribe such as Melitaeini with two subtribes as potential sister taxa, an
approach of examining all species taxa in Kallimini and Nymphalini would be
prohibitively impractical.


588
Further Description: Pattern characters of the body and appendages are identical to those
of C. perlula and C. chinatiensis with few exceptions; for the remaining characters refer
to C. chinatiensis (above). The vertex lacks orange scales as in C. perlula and unlike C.
chinatiensis. All specimens have the pattern of scales on the meso- and metathoracic
femur with orange scales dorsally and white scales ventrally, and no specimens were
found with a row of gray scales bordering the white scales as noted in a specimen of C.
perlula.
Some wing pattern characters are highly variable between and within geographic
areas and the distribution of this variation is discussed under geographic variation
(below). Forewing and hindwing fringe with the usual black and white checkered pattern
with white sections between the veins. The thin row of orange scales is present in the
basal forewing costal area.
Dorsal forewing background color varies from predominately orange to
predominately black. Median band (cells CuA2 through R3) highly variable in width and
color, varying from pale orange-cream to yellow-cream to pale whitish cream.
Postmedian band light to dark orange, usually in cells CuA2-R5 and R3 but reduced to
diffuse scaling or absent in some cells in some specimens, distinctly separate from to
broadly blended with the median band. A prominent cream patch is usually present distal
to the postmedian band in cell M3 and smaller patches sometimes present in cells CuAl
and M2, with cream patches distal to the postmedian band usually reduced to diffuse
scaling or absent in remaining cells but sometimes forming a row of distinct patches in
cells CuA2-R3. Area distal to the postmedian band black except for the aforementioned
cream patches. Discal cell with two light patches concolorous with the median band, one


342
these trees is presented in Figure 335. There was little change in resolution within
Chlosyne, but a peculiar topology was obtained for some of the taxa outside Chlosyne.
All clades obtained from the strict consensus tree from the analysis of genitalic
characters only are congruent with the analyses including all characters (Figures 311-
312). This was likewise the case for the strict consensus tree of equally weighted pattern
characters, but not for the strict consensus tree of successively weighted pattern
characters. Relationships within Chlosyne remained congruent, but some incongruent
arrangements were obtained with respect to splicing Atlantea and Higginsius into the
Chlosyniti.
A comparison between the strict consensus trees for equally weighted genitalic
and pattern characters shows that all clades are either present in both strict consensus
trees or present in one but unresolved within the other (there are no instances of
incongruence, and ten clades are common to both data sets). The same is true for
relationships within the genus Chlosyne for the strict consensus trees based on successive
weighting, however there are several cases of incongruence for relationships outside the
genus Chlosyne between the two data sets. The relationships within Microtia are
incongruent on the pattern character tree except for (comeara,anmalas). Relationships
and monophyly within Atlantea and Higginsius are congruent, however the tree for
pattern characters places Higginsius within Microtia and Atlantea as a basal taxon to
Chlosyne.
Comparing trees derived from genitalic characters and wing pattern characters
based on the UPGMA algorithm, 77% (37/48) of the groups obtained from one data set
were in conflict with the other data set (versus 0% for equally weighted characters with


682
patch. Some individuals also have small white patches of white scales touching the eyes
around the tufts of white scales ventro-lateral of the antennal bases. In some specimens
these various white markings are fused together such that the entire face may be
predominately white. Patch of orange scales at the edge of the inner anterior-lateral
margin of the eye usually absent but sometimes present but small (mostly form saundersi
and some adjutrix). Black and sometimes white scales present in the collar between the
head and pronotum. Dorsal and lateral sides of abdomen covered with black scales
without light colored bands at the edges of segments, some scattered orange scales are
also present in some individuals of forms saundersi and adjutrix. Ventrally abdomen
highly variable with respect to the extent of black scaling, confined to the typical parallel
stripes (with a cream background present) in many Chlosyne in form adjutrix and some
crocale, with an increasing amount of black moving along the continuum toward form
quehtala which has a primarily dark venter with sparse cream scaling. The amount of
cream scaling the becomes progressively greater moving along the continuum toward
form saundersi, and some saundersi have the typical black parallel stripes (mostly west
Panama specimens) but in most South American specimens of forms saundersi and
paupera the stripes are partially or completely orange instead of black. Pattern of
prothoracic legs highly variable; black and white, white and orange, or all three colors of
hairs and scales may be present and in variable patterns. Femur of meso- and
metathoracic legs with orange scales dorsally and white scales ventrally in forms
adjutrix, crocale, saundersi, and paupera but black scales ventrally in forms quehtala
and lacinia and most intermediates with these and the lighter forms. C. lacinia is the
only taxon in the Chlosyniti where the pattern on the meso- and metathoracic femur is


496
background color in C. nycteis and C. gorgone, whereas in C. harrisii the color of this
area is the same as the remaining background (at most very slightly and subtly different).
The ventral hindwing postmedian dots are surrounded by orange and no brown in C.
harrisii, while in C. nycteis these dots are surrounded by brown or by a ring of orange
scaling within brown scaling (occasionally individuals approach C. harrisii for this
character). In C. nycteis the bands between the bands of the basal symmetry system and
between the bands of the central symmetry system are brown or a diffusion of orange and
brown, whereas they are clean orange in C. harrisii. Median lines and symmetry system
bands are brown in C. nycteis and black in C. harrisii, although this character is less
helpful for worn specimens. The above characters also separate C. harrisii from C.
gorgone, in addition to the arrow head shaped sections of the distal median line and
submarginal band in C. gorgone. Other characters separating C. harrisii from these two
taxa are presented in Table 5.
Chlosyne harrisii is actually most similar in terms of ventral wing pattern to C.
kendallorum, although due to the highly disjunct ranges of these two taxa it would seem
extremely unlikely they would ever co-occur. While I found C. harrisii and C.
kendallorum can easily be separated from gestalt appearance (C. kendallorum resembles
C. gorgone dorsally), a careful examination of components of the wing pattern failed to
uncover any consistent differences. C. kendallorum seems to have a sharper pattern
between black and orange areas whereas C. harrisii is more diffused. However, the
genitalia between the two taxa are quite distinctive, granted only one representative of
each sex was examined for C. kendallorum while seven females and twelve males were
dissected and examined for C. harrisii. The male genitalia of C. harrisii have the


499
scales ventrally, tibia and tarsi covered with orange scales only. Meso- and metapleurons
where the femur tucks in with black scales and hairs and some white scales underneath,
thorax ventral to the wings densely covered with white scales and hairs with some dark
scales under the white ones. Antennal shaft with a checkered pattern of black and white
scales without orange scales, club with a white patch on the outer lateral side.
Narrow band of orange scales along the base of the forewing costal margin. Wing
fringes checkered black and white with white sections between where the veins reach the
wing margin. Basal part of ventral forewing cell C orange along the edges of the cell and
predominately white in the middle. Discal spot present with orange scaling inside.
Forewing discal cell with a patch formed from the fusion of the distal band of the basal
symmetry system and proximal band of the central symmetry system with orange scaling
inside the same color orange as that inside the discal cell. Dorsal forewing discal cell
with a single orange patch between the proximal band of the central symmetry system
and the discal spot. Basal band of basal symmetry system forming a basal patch in the
discal cell with diffuse orange scaling of identical color to the orange inside the discal
spot. Dorsal forewing cell CuA2 with the distal element of the basal symmetry system
single and forming a narrow, elongate patch with diffuse orange scaling inside extending
from the basal part of the cell almost to the origin of vein CuA2. Dorsal hindwing
postmedian band present and in the form of a series of orange patches in cells CuA2-R5,
ventrally patches in cells CuA2-Sc+Rl with orange patches in CuA2-Ml and cream
background color and a variable amount of black in cells R5-Sc+Rl (Sc+Rl may contain
a black patch). Forewing postmedian band of orange patches with cream colored
postmedian dots inside. Hindwing postmedian dots present within the postmedian band


550
-Melitaea palla vallismortis J. W. Johnson, 1938. Bull. S. California Acad. Sci. 37:18.
Type Locality: Tuber Canyon, Panamint Range, Death Valley, California.
Holotype: Los Angeles County Museum (Miller and Brown 1981).
=Chlosynepalla f. dorothyae Bauer, 1975. In Butterflies of N. America by W. H. Howe.
Type Locality: reported as Burnt and Snake River Canyons, Oregon (Miller and
Brown 1981) but actually Durkee, Baker County, Oregon (Ferris 1989).
Holotype: Los Angeles County Museum (Ferris 1989).
=Chlosyne acastus waucoba Emmel, Emmel & Mattoon, 1998. Systematics of Western
North American Butterflies: 142-143 & 155 (figures 13-16)). Type Locality:
California: Inyo County, 0.5 miles north of Cerro Gordo, Inyo Mountains, 8100'
elevation. Holotype: Natural History Museum of Los Angeles County.
=Chlosyne acastus robusta Austin, 1998. Systematics of Western North American
Butterflies: 576-577 & 583. Type Locality: Nevada: Clark County, Spring
Mountains, Kyle Canyon Campground, 2072m, T19S R 57E S28 on USGS
Charleston Peak, Nev. 15' quadrangle. Holotype: "will be deposited in the Allyn
Museum" (Austin 1998).
Diagnosis: See the diagnosis of C. palla and C. whitneyi (above).
Further Description: Genitalia identical to C. palla and C. gabbii. The pattern characters
reported in the further description of C. hoffmanni also apply to C. acastus, with the
exception of the cream background color (white to pearly white in C. acastus) and the
characteristics which distinguish C. hoffmanni from C. palla (see C. hoffmanni diagnosis
above), which also separate C. hoffmanni and C. acastus. Dark females with the dorsal


339
leveling off, which continued up to the maximum number of taxa included in the
analysis. The rescaled consistency index was most impacted by increasing the numbers
of taxa in the analysis, while the decrease was lowest for the retention index.
Variation in Character Statistics with Different Numbers of Taxa in the Analysis
Table 9 presents a mean value for the consistency index, retention index, and
rescaled consistency index for all parsimony informative characters included in each
analysis. In addition, a mean value was calculated separately for genitalic and pattern
characters. As above, node numbers corresponding to the different monophyletic groups
used in each analysis are designated on Figure 312. Graphs showing the relationship
between mean character statistics and the number of taxa in the analysis appear in Figures
329, 330, and 331 for the consistency index, retention index, and rescaled consistency
index, respectively.
A similar pattern was obtained for mean character statistics as was found for tree
statistics. Initially, when the mean value of a character statistic began to drop below 1.0
there was a sharp decrease, followed by a leveling off up to the maximum number of taxa
included in the analysis. Both genitalic and pattern characters exhibited this pattern, in
addition to the mean value of statistics for all characters combined. However, for
genitalic characters the decrease occurred after more taxa were added to the analysis than
for pattern characters, it was not as steep, and the leveling off did not occur until slightly
more taxa had been added to the analysis relative to pattern characters. Also, the values
around which mean genitalic character statistics began to level off were notably higher
than for pattern characters. Similar patterns were found for mean consistency index,
mean retention index, and mean rescaled consistency index.


394
Table 5
Continued
Characters
Taxa
121
122
123
124
125
126
127
128
129
130
131 I
132
133
134
) 135
136
137
138
139 l
140
141 ¡ 142
143
144
Chlosyne nycteis
0
0
0
0
0
0
?
1
0
0
0
0
1
1
0,4
0
0
1
1
1
0 0
0
0
Chlosyne gorgone
0
0
0
0
0
0
?
1
0
0
0
0
1
1
0,4
0
0
1
1
1
0 0
0
0
Chlosyne harrissii
0
0
0
0
0
0
?
1
0
0
0
0
1
0
0
0
0
0
0
0
0 0
0
0
Chlosyne kendallorum
0
0
0
0
0
0
?
1
0
0
0
0
1
0
0
0
0
0
0
0
0 0
0
0
Chlosyne hoffmanni
0
0
0
0
0
0
?
0
0
0
0
0
1
0
0
0
0
0
0
0
0 0
0
0
Chlosyne palla
0
0
0
0
0
0
?
0
0
0
0
0
1
0
0
0
0
0
0
0
0 0
0
0
Chlosyne gabbii
0
0
0
0
0
0
?
0
0
0
0
0
1
0
0
0
0
0
0
0
0 0
0
0
Chlosyne acastus
0
0
0 0
0
0
?
0
0
0
0
0
1
0
0
0
0
0
0
0
0 0
0
0
Chlosyne whitneyi damoetas
0
0
0 0
0
0
?
0
0
0
0
0
1
0
0
0
0
0
0
0
0 0
0
0
Chlosyne whitneyi whitneyi
0
0
0
0
0
0
?
0
0
0
0
0
1
0
0
0
0
0
0
0
0 0
0
0
Chlosyne definita
1
0
0
0
0
0
?
5
0
0
0
0
1
0
0
0
0
0
0
0
0 0
0
0
Thessalia ezra
1
0
0
0
0
0
?
3
0
0,3
0
*
2
*
?
0.1
0
0
3
0
1 1
2
0
Thessalia theona
1
0
0
0
0
0
?
3
0
0
0
0
2
*
?
0,1
0
0
0
0
1 1
1
0
Thessalia perlula
1
4
0
0
0
0
?
3
0
0
0
0
2
*
?
0.1
0
0
0
0
1 1
1
0
Thessalia chinatiensis
1
0
0
0
0
0
?
3
0
0
0
*
2
*
?
0.1
0
0
0
0
1 1
1
0
Thessalia leanira leanira
1
0
0
0
0
0
?
3,4
2
0.3
0
0
2
*
?
0.2
0
?
?
0
0 1
2
0
Thessalia leanira fulvia
1
0
0
0
0
0
?
4
2
0
0
0
2
*
?
2
?
?
0
0 1
2
0
Thessalia cyneas
1
0
0
0
0
0
?
4
2
0
0
0
2
*
?
2
?
*
?
0
0 1
2
0
Thessalia cynisca
1
0
0
0
0
0
?
3
2
3
0
*
2
*
?
2
?
*
?
?
0 1
2
0
Chlosyne endeis pardelina
1
0
0
0
0
0
?
3
0
0
0
0
1
0
2
0
0
0
0.2
0
0 0
0
0
Chlosyne endeis endeis
1
0
0
0
0
0
?
3
o
0
0
0
1
0
2
0
0
0
0.2
0
0 0
0
0
Chlosyne melitaeioides
1
0
0
0
0
0
?
3
0
?
0
0
1
0
1,2
0
1
0
3
0
0 0
0
0
Chlosyne marina
1
0
0
0
0
0
?
3
0
1
0
0
1
0
1.2
0
1
0
3
0
0 0
0
0
Chlosyne poecile
1
0
0
0
0
0
?
3
0
1
0
0
1
o
1
0
1
0
3
0
0 0
0
0
Chlosyne erodyle
1
3
0
0
0
0
?
3
0
1
0
0
1
0
1
0
1
0
3
0
0 0
0
0
Chlosyne melanarge
1
*
0
0
0
0
?
2,3
0
1
0
*
?
?
?
?
?
?
3
?
0 0
0
*
Chlosyne hylaeus
1
0
0
0
0
0
?
?
0
1
0
0
1
0
1
0
1
0
3
0
0 0
0
0
Chlosyne eumeda
1
0
0
0
0
0
?
2
0
1
0
0
1
0
1
0
1
0
3
0
0 0
0
0
Chlosyne californica
1
0
2
0
0
?
?
6
0
2
0
0
1
0
1
0
0
0
3
0
0 0
0
0
Chlosyne lacinia
1
0
2
0
0
?
?
?
0
1
0
0
1
0
1
0
0
0
3
0
0 0
0
0
Chlosyne ehrenbergii
*
*
0
0
0
0
?
7
2
1
0
*
2
*
?
2
?
*
?
?
1 1
2
0
Chlosyne hippodrome
1
*
1
1
?
?
?
2
0
1
0
*
?
?
?
?
?
?
?
?
0 0
0
0
Chlosyne narva
1
0
0
0
0
0
?
2
0
1
0
*
1
0
1
3
0
0
3
0
0 0
0
0
Chlosyne gaudealis
1
*
0
0
0
0
?
2
0
1
0
*
1
0
1
3
0
0
3
0
0 0
0
0
C. gaudealis wellingi
1
*
0
0
0
0
?
2
0
1
0
*
?
?
?
3
?
0
?
?
0 0
0
0
Chlosyne janais
1
1
1
2
2
1
1
2
0
1
0
*
1
0
1
0
1
0
3
0
0 0
0
0
C. marianna
1
*
1
2
2
1,2
?
2
0
1
0
*
?
?
?
?
?
0
?
?
0 0
0
0
Chlosyne rosita rosita
1
1
1
3
1
1
1
2
1
*
0
*
1
0
1
0
1
0
3
0
0 0
0
0
Chlosyne montana
1
1
1
3
1
1
2
2
1
*
0
*
1
0
1
0
1
0
3
0
0 0
0
0
Chlosyne rosita browni
1
1
1
3
1
1
2
2
1
*
0
*
1
0
1
0
1
0
3
0
0 0
0
0
Chlosyne riobalensis
1
*
1
3
1
2
?
2
1
*
0
*
?
?
?
?
?
?
?
?
0 0
0
*
Chlosyne mazarum
1
2
1
3
1
2
3
2
1
*
0
*
1
0
1
0
1
0
3
0
0 0
0
0
Texola elada
0
0
0
0
0
0
?
0
0
0
0
0
0
0
0
0
0
0
0
0
0 0
0
0
Texola coracara
0
*
0
0
0
0
?
3
0
0
0
*
0
0
0
0
0
0
0
0
0 0
0
0
Texola anomalus
0
1
0
0
0
0
?
3
0
0
0
*
0
0
0
0
0
0
0
0
0 0
0
0
Dymasia dymas
0
0
0
0
0
0
?
0
0
0
0
0
0
0
0
0
0
0
0
0
0 0
0
0
Microtia elva
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
? ?
?
?
Antillea pelops
0
0
0
0
0
0
?
0
0
?
1
0
0
2
3
0
0
3
4
0
0 0
0
1
Antillea proclea
0
0
0
0
0
0
?
0
0
?
1
0
0
2
3
0
0
3
4
0
0 0
0
1
Poladryas minuta
0
0
0
0
0
0
?
0
0
0
0
0
0
0
0
0
0
0
0
0
0 0
0
0
Higginsius fasciatus
0
0
0
0
0
0
?
8
0
4
0
1
0
3
0
0
0
4
0
0
0 0
0
0
Higginsius miriam
0
0
0
0
0
0
?
8
0
4.5
0
1
0
3
0
0
0
4
0
0
0 0
0
0
Atlantea perezi
0
0
0
0
0
0
?
1
3
0
0
0
?
0
1
?
0
?
3
0
0 0
0
0
Atlantea tulita
0
0
0
0
0
0
?
1
3
0
0
0
0
?
1
0
0
0
3
0
0 0
0
0
Atlantea pantoni
0
0
0
0
0
0
?
1
3
0
0
0
0
0
1
0
0
0
3
0
0 0
0
0
?
Not coded (see character descriptions for explanation
)
*
Indicates a taxon lacks a primitive character that is subdivided into forms as a separate
:haracter (=?)
N
No material examined i
=?)


35
information within the Euphydryiti. For the following character, I code taxa lacking a
forked inner valve process as
11. For those taxa with a forked inner valve process, the orientation of the tip of the
ventral fork:
l=Curved ventrally outward (Figs. 7-8).
2=Curved ventrally inward (Figs. 3 & 9-14).
3=Strongly inward and slightly ventrally (Fig. 6).
The variation within taxa with a saw-like toothed inner valve process (character 4
state 4) can be further subdivided into discrete states. Independent character evidence
suggests that this form of the inner valve process is a terminal derived state.
12. For those taxa with a saw-like inner valve process, the distal extent of the teeth on
the process:
l=Distal half free of teeth.
2=Teeth continue distinctly distal of the distal half (Fig. 34).
Taxa lacking a saw-like inner valve process are coded 0.
13. For those taxa with a inner valve process (character 1 state 2), the shape of distal end
of this process:
l=Narrow and sharply or bluntly pointed (Fig. 1).
2=Flared out and serrate (Figs. 10-14).
Most state 1 taxa examined have a sharply pointed process; however, it is bluntly pointed
in some derived Chlosyne. This variation is addressed in the data matrix for the
Chlosyniti.


636
Locality & Data: "La Gloria, S. of Monclova, Coah. Mex., 3300 feet, Aug. 24th,
1947". Holotype: American Museum, New York (Higgins 1960).
Diagnosis: Like C. emleis endeis except for the differences noted in the diagnosis of
subspecies endeis above.
Further Description: The color of the dorsal forewing median band is sexually
dimorphic, being yellowish cream in males and white in females.
Range: I have seen specimens from three U.S. counties, all in southern Texas (Duval,
Hidalgo, and Starr)), and from the Mexican states of Coahuila, Hidalgo, San Luis Potosi,
and Tamaulipas. Table 12 includes data from the NMNH, FSCA, and AM.
Chlosyne marina Geyer
Figure 288
Araschnia marina Geyer, 1837. Zutrage. 5:20. Type Locality: Mexico.
Note: Higgins (1960) reported the holotype of C. marina to be in the British Museum,
but according to Gerardo Lamas (pers. com. 2000), it is not there and is probably lost. I
have been able to examine the types of C. melitaeoides, C. eumeda, and C. hylaeus,
courtesy of photos sent by Gerardo Lamas.
Diagnosis: Chlosyne marina can be confused with C. endeis endeis, and may be
separated by the characters given in the diagnosis of C. endeis endeis. In collections I
have also found it confused with C. erodyle erodyle, C. melitaeoides, and C. eumeda.


305
Chlosyne cynisca is coded "?" for the same reason as for character 99.
Among those taxa with state 1, C. gorgone, C. nycteis, and C. harrissii always
have prominent dots in cells CuA2-Ml and usually R5. C. palla, C. hoffmani, C. gabbi,
and C. damaetas are intraspecifically variable to how many hindwing cells are occupied
by postmedian dots. Polydryas minuta is polymorphic for the presence or absence of
postmedian dots, but actually variation seems to occur along a continuum between
prominent dots, very small ones, and none at all. I code this taxon 0&l.The specimens
of Higginsius miriam examined have prominent upperside dots in CuA2-R5. However,
two of the five specimens of H.fasciatus (all males) examined have the postmedian dots
incomplete and only conspicuous in cell CuA2. Of the four specimens of Chlosyne
perlula examined, one had a diffuse row of black postmedian dots. Only two of the long
series of C. theona examined had a similar row of diffuse black postmedian dots. These
taxa are coded 0&6.
Among state 2 taxa, C. narva seems to consistently have prominent dots in CuA2-
R5 while the other taxa vary greatly intraspecifically between this condition to very
obscure or even no dots on the upper wing surface (but postmedian dots are present on
the under wing surface in these specimens), or their variation occupies a portion of this
continuum. State 5 and state 7 taxa have the postmedian dots occupying cells CuA2-R5.
111. For those taxa with hindwing postmedian dots, comparison of the hindwing
postmedian dots on the dorsal and ventral wing surfaces.
l=Similar on upper and under wing surfaces but with under surface dots usually
more prominent (never less prominent).


52
l=Absent.
2=Present (Higgins 1981, Fig. 475 pg. 236).
48. A heavily sclerotized convexly curved claw on each posterior side of the uncus:
l=None.
2=An entire claw is present (Figs. 74 & 76).
3=A dentate claw is present (Higgins 1981, Fig. 473 pg. 235).
An alternative coding scheme would be to code for presence or absence of a claw, and
then have a separate character for whether the claw is smooth or dentate. However, the
structure of the dentate and entire claws and their exact position is quite different, so the
above coding seemed most appropriate. It also may be appropriate to combine the two
preceding characters as a single multistate character. These character states are particular
to the Phycioditi, and not informative other than for relationships within the Phycioditi,
and consequently extraneous to the goals of this analysis. Examination of more
Phycioditi taxa would give more insight into the range of variation occurring within
Phycioditi and the best way to code it.
49. Hollow, paired, pointed posterior projections formed from invaginations of the inner
wall of the tegumen (some of the out-group states refer to single projections-only paired
projections occur in the Melitaeini):
0=None.
l=Projections divide the tegumen into a deep fork; they are connected by only a
very thin bridge of tegumen between them and anteriorly (Figs. 4,16-23 & 25).


54
#D=A single elongate centered projection widened posteriorly with two lateral
pointed extensions.
Due to the notably different structure/characteristics and position between the
invaginations of the above in group states, I did not code for the presence or absence of
paired invaginated projections of the tegumen.
50. For those taxa with character 49 state 1, the shape of the tegumen projections:
l=Elongate, narrow, and curving ventrally to a point distally (Figs. 23-24).
2=Short, broad, and triangular with the outer side slanted and the inner side nearly
vertical (Fig. 21).
3=Prominent and triangular with a broad base and a narrow straight tip (Fig. 16).
4=Base broadly triangular but the distal portion narrower and slanting inward
toward the midline; the distal portion is blunt and more finger-like than triangular
(Fig. 19).
5=Long and pointed with the outer side concave and the inner side curving
outward distally while being more vertical/concave basally (Figs. 18,20,22 & 25).
6=Prominent, proximal approximately Vi broad and triangular, distal V2 narrower
and triangular curving sharply outward at the distal extremity (Fig. 4).
7=Short and broad, basal Vi somewhat rectangular, distal Vi projects from lateral
side of base, triangular, and with the inner side concave and outer side fairly
straight (Fig. 17).
Independent evidence from other characters suggests state 1 of the preceding character is
a terminal derived state. Consequently, taxa lacking this state are coded 0.


613
figures (black and white dorsal and ventral photos of one specimen of each sex), as
opposed to taxa of any rank or the range of variation which Austin and Smith (1998b)
included in their subspecies concepts. I do wish to comment that I find almost no
distinction between the alma, flavodorsalis, and basiensis phenotypes, and all three
appear to me to be represented in a series from Pyramid Lake, Nevada. For future
reference, I collectively refer to these three phenotypes as the alma phenotype. Also, the
austrima phenotype, limited on Austin and Smith's (1998a) range map to Baja California,
as best as I can tell from their black and white photographs and description appears to me
to be within the range of variation found in southern California specimens.
The most extreme phenotype on one end of the continuum is a phenotype with no
orange markings on the dorsal surface with just cream patches against a dark background.
The median and postmedian bands and the light patches in the basal area of the wings
contrast sharply against the black background. There is little difference in the shape of
the wings between the sexes, although females tend to be distinctly larger. The darkest
phenotype apparently occurs only along the west coast of California. Series of specimens
show a transition between this extreme dark phenotype to one with the same shaped
wings and color light markings but where almost all of the dorsal black is replaced by
orange. Moving along the transition series from dark to light, the orange first becomes
prominent dorsal forewing cells M3 to Ml distal to the postmedian dots, followed by the
one or more of the following three areas: adjacent to the basal forewing costal area,
between the forewing median and postmedian bands in the same cells as above, and distal
to the postmedian dots on the ventral hindwing. The next area to become orange after the
preceding three are filled is the forewing discal cell. The last areas to change to orange


47
2=Somewhat more heavily sclerotized than the remainder of the juxta, however
not nearly to the degree as state 1 and clearly not appearing almost black (Figs. 2
fe 5).
Since this character is scored by the sclerotization of the juxta relative to that of the
projection in the same specimen, differences in how long different specimens were left in
KOH should not prevent the ability to accurately score this character. However, the
specimens scored for this character were placed in KOH from the same stock solution at
the same time, and left in KOH at the same temperature (room temperature) and then
dissected in succession (except for Hypodryas iduna). However, I found other dissected
specimens left in KOH for differing amounts of time could still be coded the same way
without ambiguity.
Taxa lacking character 18 state 2 are coded 0, with the exception of Eurodryas
desfontaini. Many other independent characters support the monophyly of the
Euphydryiti clade, and E. desfontaini is the only Euphydryiti examined which lacks the
derived state of the preceding character. I avoid coding this taxon as 0 for the above
character, because the absence of paired juxta projections is the only character suggesting
E. desfontaini may be more primitive than the other Euphydryiti examined. Coding E.
desfontaini as 0 rather than ? would produce an unwarranted bias.
38. For those taxa with paired posterior juxta projections, the lateral compression of
these projections:
l=Distinctly laterally compressed (Fig. 5).
2=Not laterally compressed (Fig. 2).


397
Table 8: Status of individual characters in the Chlosyniti/Poladryiti data matrix prior to
successive weighting with the distinct state model for polymorphisms
Character
Weight
States
Steps
Cl
R,
RC
Character
Weight States
Steps
Cl
Rl
RC
1
1
2
1
1
1
1
73
1
8
8
0.88
0.86
0 75
2
1
4
4
0.75
0.88
0 66
74
1
3
5
0.4
0.86
0 35
3
1
2
1
1
1
1
75
1
8
10
0.7
0.84
0.59
4
1
3
2
1
1
1
76
1
5
7
0.57
0.75
0.43
5
1
3
2
1
1
1
77
1
5
7
0.57
0.75
0.43
6
1
3
2
1
1
1
78
1
6
8
063
0.7
0 44
7
1
3
2
1
1
1
79
1
11
16
063
076
0.47
8
1
2
1
1
1
1
80
1
3
5
0.6
0.87
0 52
9
1
5
4
1
1
1
81
1
2
1
1
0

10
1
3
5
0.4
0.82
0.33
82
1
3
3
0.67
0.75
0.5
11
1
3
2
1
0
0
83
1
6
5
1
1
1
12
1
4
3
1
1
1
84
1
5
4
1
1
1
13
1
5
4
1
1
1
85
1
10
13
0.69
0.88
0 61
14
1
4
4
0.75
0.8
0.6
86
1
2
3
0.33
0.78
0.26
15
1
7
6
1
1
1
87
1
3
5
0.4
0.7
028
16
1
3
2
1
1
1
88
1
4
5
0.6
0.78
0.47
17
1
3
2
1
1
1
89
1
4
6
0.5
0.25
0.13
18
1
3
2
1
1
1
90
1
6
7
0.71
0.9
0 65
19
1
4
3
1
1
1
91
1
8
10
0.7
0.88
061
20
1
4
3
1
1
1
92
1
3
3
0.67
0.83
0 56
21
1
3
2
1
1
1
93
1
7
6
1
1
1
22
1
2
1
1
1
1
94
1
3
5
0.6
0.75
0 45
23
1
5
7
0.57
0.84
0.48
95
1
2
1
1
1
1
24
1
2
2
0.5
0.67
0.33
96
1
2
2
0.5
0.75
0.38
25
1
2
1
1
1
1
97
1
3
2
1
1
1
26
1
2
1
1
1
1
98
1
3
2
1
1
1
27
1
2
1
1
1
1
99
1
3
3
067
0.92
0.62
28
1
2
1
1
1
1
100
1
3
2
1
1
1
29
1
2
2
0.5
0.67
0.33
101
1
9
9
0.89
0.86
0.76
30
1
2
1
1
1
1
102
1
6
6
0.83
0.89
0.74
31
1
2
1
1
1
1
103
1
7
7
0.86
0.97
0.83
32
1
2
1
1
1
1
104
1
2
1
1
1
1
33
1
3
2
1
1
1
105
1
3
2
1
0
0
34
1
7
6
1
1
1
106
1
8
7
1
1
1
35
1
2
1
1
1
1
107
1
5
7
0.57
0.57
0 33
36
1
2
2
1
1
1
108
1
3
7
0.29
0.79
0 23
37
1
4
3
1
1
1
109
1
2
1
1
1
1
38
1
2
1
1
1
1
110
1
11
14
0.71
0.86
0 61
39
1
3
2
1
1
1
111
1
3
2
1
1
1
40
1
2
2
0.5
0.67
0.33
112
1
4
4
0.75
0.9
0.68
41
1
3
2
1
1
1
113
1
4
3
1
1
1
42
1
2
1
1
1
1
114
1
2
1
1
1
1
43
1
2
2
1 _
1
1
115
1
4
4
0 75
0.94
0.7
44
1
2
2
1
1
1
116
1
5
7
0.57
0.79
0.45
45
1
4
3
1
1
1
117
1
5
5
08
0.91
0.73
46
1
2
1
1
1
1
118
1
9
8
1
1
1
47
1
2
1
1
1
1
119
1
2
1
1
1
1
48
1
3
2
1
1
1
120
1
8
7
1
1
1
49
1
3
2
1
1
1
121
1
2
1
1
1
1
50
1
2
1
1
1
1
122
1
5
5
0.8
0.75
0 6
51
1
3
2
1
1
1
123
1
3
3
0.67
0 88
0.58
52
1
2
1
1
1
1
124
1
4
3
1
1
1
53
1
3
6
0.5
088
0.44
125
1
3
3
0.67
0.83
0 56
54
1
1
0
0
0
0
126
1
4
3
1
1
1
55
1
7
8
0.75
0.88
0.66
127
1
3
2
1
1
1
56
1
9
12
0.67
0.76
0.51
128
1
it
11
091
0.97
0.88
57
1
10
13
0.69
0.83
0.58
129
1
4
4
0.75
0.9
0 68
58
1
10
13
0.69
0.75
0.52
130
1
7
7
0 86
0.93
0.8
59
1
2
2
0.5
0.95
0.47
131
1
2
1
1
1
1
60
1 ..
4
7
0.43
0.82
0.35
132
1
2
1
1
1
1
61
1
6
6
0.83
0 96
0.8
133
1
3
3
0.67
0.94
0.63
62
1
5
6
0 67
0.92
0 61
134
1
4
3
1
1
1
63
1
2
2
1
1
1
135
1
6
6
083
0.94
0 79
64
1
3
5
04
0.87
0.35
136
1
5
5
0.8
0.88
0.7
65
1
3
6
0 33
083
0.28
137
1
2
2
0 5
089
0 44
66
1
7
8
0.75
0.89
0.67
138
1
4
3
1
1
1
67
1
4
6
0.5
0.87
0.43
139
1
5
5
0.8
0 95
0.76
68
1
5
7
0.57
0.88
0.5
140
1
2
1
1
1
1
69
1
5
8
0.5
0.75
0 38
141
1
2
2
0.5
0.75
0 38
70
1
2
1
1
0
0
142
1
2
2
0.5
0 88
0.44
71
1
3
6
0.33
0.85
0.28
143
1
3
3
067
0 86
0.57
72
1
9
8
1
1
1
144
1
2
1
1
1
1


601
Further Description: Several pattern characters of the body and appendages vary between
and within subspecies. Palpi variable between subspecies. Vertex with a distinct white
centered patch reaching the posterior edge of the plate; black scales always present but
the presence of orange scales on the vertex varies between and within subspecies.
Sutures at the lateral edges of the vertex always with black and white scales, sometimes
with orange scales. Tuft of white hairs on the frontoclypeus anterior-lateral of each
antennal base well developed. Frontoclypeus pattern varies between subspecies. Patch
of orange scales present at the edge of the inner anterior-lateral margin of the eye. Black
and white scales present in the collar between the head and pronotum, presence of orange
scales varies between subspecies. Posterior edges of posterior abdominal segments with
bands of white scales on dorsal and lateral sides. Ventrally abdomen with two
longitudinal parallel black stripes against a pale to yellowish cream background, usually
not appearing broken between each segment. Tibia and tarsi of prothoracic legs with
orange scales dorsally and white scales and hairs ventrally. Femur of meso- and
metathoracic legs with predominately orange scales dorsally and white scales ventrally
except at distal end where only orange scales are present (white scales lacking in some
specimens of subspecies leanira), tibia and tarsi with orange scales only. Meso- and
metapleurons where the femur tucks in with black scales and hairs, with white scales
sometimes visible underneath the dark ones. Thorax ventral to the wings with dense
white and pale orange with some black scales underneath the hairs; subspecies leanira
may have the white hairs replaced by yellow tinged cream hairs in the western part of its
range, and may lack the orange hairs. Antennal shaft variably checkered black and white,
with orange scales present or absent. Club with a white patch on the outer lateral side.


325
complete) encompasses part of cell CuA2 (distal to where vein CuAl branches off) and
cell 1A+2A, where both the basal and distal bands of the central symmetry system are
again present and forming the borders of the band. The sizes of the components of the
band in the various cells is an intraspecifically variable feature.
There is substantial variation in the form of this band in many taxa which is
different from the variation in the band formed between the bands of the basal symmetry
system, hence this band warrants coding as additional characters. However, although the
components in each cell forming this band appear to be more coalesced in Chlosyne than
in the other genera (although the difference is not as pronounced as that between the band
and series of patches formed between the bands of the basal symmetry system) 1 do not
code this variation because I believe it would duplicate character (17) (all taxa would be
coded the same for states 0 vs. 1 suggesting dependence).
Note that in some individuals of some taxa portions of this band are visible on the
dorsal wing surfaces as well. However, when this occurs the pattern is like that on the
ventral surface except obscured by dark scaling in the basal area. This character is best
coded based exclusively on the ventral surface, as there is no new information from the
dorsal surface and the pattern is largely or completely obscured there.
136. Presence and completeness of a band formed between the proximal and distal
elements of the central symmetry system on the ventral hindwings as described above:
0=Present and complete.
l=The sections in cells Sc+Rl through the discal cell are present, but the
remainder is absent.
2=Completely absent.


412
Bootstrap
99
56
73
63
85
79
94
51
88
79
59
90
83
54
80
98
81
61
77
66
96
64
51
65
91
84
79
92
55
60
60
100
100
60
100
56
84
80
- C. nycteis
C. gorgone
- C. harrissii
- C. n. sp. nr. har.
- C. hoffmanni
- C. palla
C. gabbii
- C. acastus
- C. w. damoetas
C. w. whitneyi
C. definita
T. ezra
T. theona
T. perlula
- T. chinatiensis
T. I. leanira
T. I. fulvia
T.cyneas
T. cynisca
C. e. pardelina
C. e. endeis
C. melitaeoides
C. marina
C. e. poecile
C. e. erodyle
C. melanarge
C. eumeda
C. californica
C. lacinia
C. ehrenbergi
C. hippodrome
C. narva
C. g. gaudealis
C. g. wellingi
C. janais
C. marianna
C. rosita rosita
C. montana
C. rosita browni
C. riobalensis
C. mazarum
T. eleda
D. dymas
M. elva
T. coracara
T. anomalus
A. pelops
A. proclea
P. minuta
H. fasciatus
H. miriam
A. perezi
A. pantoni
A. tulita
CumPhyGnaMel
Figure 317: Boot strap 50% consensus tree for equally weighted characters from the Chlosyniti/
Poladryiti data matrix, with multistate taxa assigned discrete states.


368
In my view, I am highly skeptical that boot strap scores actually provide much
indication for the support of particular clades, or even their relative support compared
with other clades derived from the same data set. In my case study, there is no consistent
pattern for boot strap scores as the size of a monophyletic in group increases. Particular
clades have boot strap scores which increase, decrease, stay fairly constant, or fluctuate
as more taxa are added to the analysis, and consequently the range between boot strap
scores of particular clades in data sets with different numbers of taxa is highly variable
(Figure 336). Consequently, in contrast to tree statistics (Cl, RI, and RC) which have a
general pattern of decrease as more taxa are added to the data set, boot strap scores seem
to have no general pattern with respect to adding taxa other than that they are highly
variable. For example, node 34 has a boot strap score of 47% compared with 96% for
node 44 when a monophyletic group of 30 taxa is included in the analysis (see Figure 311
for clades associated with node numbers); however, when a monophyletic group of 18
taxa is included in the analysis, node 34 has a score of 97% compared with 95% for node
44. With 41 taxa in the analysis, node 16 has a boot strap score of 39% compared with
96% for node 20; however, with 54 taxa in the analysis node 16 and node 20 have boot
strap scores of 78% and 77%, respectively. Nodes 36 and 43 have boot strap scores of
94+/- 4% for any number of taxa in the analysis between 14 and 54. I argue that the high
variability with respect to absolute and relative values of boot strap scores with different
sized monophyletic groups included in the analysis is inconsistent with the hypothesis
that boot strap scores provide an indication of support for particular clades. Boot strap
scores appear in many cases to be highly incidental to the size of the monophyletic group
in the analysis.


684
and R3, with the CuA2 dot sometimes split into two by dark scaling along the veinlet.
Ventral hindwing median lines are absent based on specimens with a light background
color. Area between the positions of the submarginal and marginal bands (neither band is
detectable) solid black dorsally and ventrally on forewings and hindwings. Basal and
distal elements of the both the basal and central symmetry systems black with solid black
in-between, forming a solid black band for each symmetry system detectable in
individuals with a light ventral hindwing background. Individuals with a light ventral
background lack black scaling along the veins for much of their length and have only
very thin black scaling along veins distally.
Range: C. lacinia is the most widely distributed member of the Chlosyniti. In the United
States I have seen material from California, Nevada, Arizona, Utah, New Mexico, Texas,
and Oklahoma. According to Scott (1986) strays have been recorded as far north as
Minnesota, and Opler and Krizek (1984) report strays from Barry and Jackson Counties
in western Missouri. It appears to occur throughout all of Mexico, and I have seen
records from many of the Mexican states except for the northwest part of the country
including Sonora and Baja California. However, Scott (1986) has all of Sonora
highlighted in his range map, and Brown et. al. (1992) show two distributional dots from
northeast Baja California Norte near the U.S. border and two distributional dots from
southern Baja California Sur. Likewise, it is present throughout Central America
although I have seen no specimens from Belize, although I expect it does occur there. C.
lacinia is one of the few Chlosyniti thats range reaches South America, and it is the only
representative with an extensive distribution on that continent. I have seen material from
Trinidad, Venezuela, Colombia, Ecuador, Peru, Bolivia, Paraguay, northern Argentina,


261
2=0nly black scales and hairs present.
*3=Sexually dimorphic, orange and white scales in female and only white scales
in male.
@*A=1&2: Individuals are either state 1 or state 2.
Many taxa are intraspecifically variable in having either state 1 or state 2, and are coded
1&2. This variation may occur within both sexes. Only one subspecies was found to
be sexually dimorphic, so this character was not coded separately for males and females,
since this would have weighted most of the state changes twice.
Abdomen:
All taxa have black scales on the abdomen, but the presence or absence of white
and orange scales and their patterns varies among taxa. The presence and characteristics
of light colored bands around the posterior edge of abdominal segments also varies
among taxa.
69. Presence of thin bands (may be broken) of light-colored scales around the dorsal and
lateral posterior edges of the posterior (and sometimes all) abdominal segments.
0=White bands present.
l=Light colored bands of scales absent.
2=Orange bands present.
#3=Anterior three segments have a white band dorsally, segments posterior to
these have an orange band dorsally (some white scales may be mixed in), laterally
(and ventrally) the orange bands become mixed with white and change to white
ventrally; the extent of the transition to white decreases with each segment
moving anteriorly to posteriorly.


31
3. For those taxa with character 1 state 2, the curve of the posterior side of the inner
valve process in ventral view:
l=Concave (Fig. 37).
2=Convex (Fig. 108).
3=Both sides fairly straight, neither distinctly concave or convex(Figs. 89 & 87).
The exact angle from which the male genitalia are viewed affects the apparent shape of
the inner valve process. In ventral aspect, as the anterior end of the genitalia are pushed
down (increasing the posterior aspect of the view) the posterior edge begins to look
concave for both state 1 and state 2 taxa. The state differences appear to be due to
differences in how the base of the inner valve process has been rotated. Within the genus
Chlosyne both states 1 and 2 occur, in addition to an intermediate state which occurs in
four taxa, which does not affect this analysis. One taxon (C. leanira and its subspecies)
exhibits integrades between state 2 and the intermediate state, supporting the rotation
hypothesis. However, Eurodryas desfontaini appears to have independently acquired a
state 2 process due to a change in shape rather than rotation, as this taxa has a forked
inner valve process with the dorsal fork is in the same position as other Euphydryiti with
state 1.
Since Gnathotriche sodialis and G. exclamationis have the inner valve process
hinged down, the frame of reference for these taxa is different. However, if I am correct
that this variation is due to the base of the process being hinged down, these taxa would
be state 1. To avoid making this additional assumption, I code these taxa as ? along
with G. mundina which has a vestigial inner valve process. The out group is coded the
same as for the preceding character.


398
Table
9: Vari
tion in tree statistics
and the proportion of homoplastic characters with differing numbers of taxa included in the analysis
Tree
Tree
Tree
SW Tree
Genitalia Characters
Pattern Characters
All Characters
No. Parsimony
Informative Chars
No. Homoplastic
Characters
Proportion of
No.
SW Tree
SW Tree
Mean
Mean
Mean
Mean
Mean
Mean
Mean
Mean
Mean
Chars
Homoplastic
Node**
Taxa
Cl
Rl
RC
Cl
Rl
RC
Cl
0.967
RC
0.924
Cl
Rl
RC
Cl
Rl
RC
Genitalia
Pattern
Total
Genitalia
Pattern
Total
Genitalia
Pattern
Total
1 54
0.796
0.898
0.715
0.858
0.935
0.802
0.944
0.790
0.889
0.717
0.846
0.918
0.793
51
88
139
7
58
65
0.137
0.659
0.468
4 48
0.813
0.899
0.731
0.866
0.936
0.811
0.954
0.971
0.936
0.813
0.899
0.747
0.855
0.920
0.803
36
86
122
4
49
53
0.111
0.570
0.434
7 46
0.803
0.895
0.719
0.854
0.933
0.798
0.950
0.966
0.931
0.800
0.887
0.727
0.836
0.906
0.775
25
80
105
3
48
51
0.120
0.600
0.486
8 41
0.845
0.918
0.776
0.889
0.951
0.846
0.921
0.950
0.893
0.831
0.913
0.775
0.846
0.919
0.795
16
78
94
3
40
43
0.188
0.513
0.457
12 31
0.833
0.882
0.735
0.876
0.928
0.813
0.967
0.995
0.964
0.786
0.859
0.700
0.820
0.885
0.751
15
64
79
1
38
39
0.067
0.594
0.494
iS 30-
0.845
0.891
0.753
0.884
0.934 0.825
0.967
0.995
0.964
0.800
0.862
0.724
0.833
0.888
0.764
15
63
78
1
36
37
0.067
0.571
0.474
17 22
0.919
0.917
0.843
0.960
0.970
0.932
1.000
1.000
1.000
0.910
0.926
0.863
0.931
0.945
0.896
14
45
59
0
11
11
0.000
0.244
0.186
18 20
0.907
0.901
0.817
0.960
0.969
0.930
1.000
1.000
1.000
0.897
0.908
0.835
0.926
0.934
0.881
14
35
49
0
11
11
0.000
0.314
0.224
19 18
0.901
0.894
0.805
0.966
0.971
0.938
1.000
1.000
1.000
0.892
0.915
0.837
0.9251 0.941
0.887
14
31
45
0
9
9
0.000
0.290
0.200
20 15
0.932
0.920
0.857
0.974
0.973
0.948
1.000
1.000
1.000
0.919
0.944
0.881
0.946
0.963
0.921
13
26
39
0
6
6
0.000
0.231
0.154
21 i 14
0.939
0.931
0.875
0.965
0.967
0.933
1.000
1.000
1.000
0.906
0.933
0.863
0.940
0.957
0.913
13
23
36
0
6
6
0.000
0.261
0.167
9 10
0.984
0.976
0.960
0.991
0.991
0.982
0.833
0.890
0.777
1.000
1.000
1.000
0.975
0.984
0.967
3
17
20
1
0
1
0.333
0.000
0.050
14 8
0.923
0.870
0.803
0.969
0.971
0.942
1.000
1.000
1.000
0.948
0.948
0.916
0.954
0.954
0.926
2
16
18
0
2
2
0.000
0.125
0.111
22 7
0.973
0.913
0.888
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
9
9
18
0
0
0
0.000
0.000
0.000
25 7
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1
10
11
0
0
0
0.000
0.000
0.000
2
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
9
37
46
0
0
0
0.000
0.000
0.000
TTl 6
1.000
1.000
1.000
1.000
1.000
1.000
*
*
*
1.000
1.000
1.000
1.000
1.000
1.000
0
1
1
0
0
0
*
0.000
0.000
3 5
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
8
33
41
0
0
0
0.000
0.000
0.000
5 5
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
4
14
18
0
0
0
0.000
0.000
0.000
23 5
1.000
1.000
1.000
1.000
1.000
1.000
1.000
i.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
2
6
8
0
0
0
0.000
0.000
0.000
26 5
1.000
1.000
1.000
1.000
1.000
1.000
*
*
*
1.000
1.000
1.000
1.000
1.000
1.000
0
1
1
0
0
0
*
0.000
0.000
6 4
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
2
4
6
0
0
0
0.000
0.000
0.000
10 4
1.000
1.000
1.000
1.000
1.000
1.000
*
*
*
1.000
1.000
1.000
1.000
1.000
1.000
0
8
8
0
0
0
*
0.000
0.000
24 4
1.000
1.000
1.000
1.000
1.000
1.000
*
*
*
1.000
1.000
1.000
1.000
1.000
1.000
0
4
4
0
0
0
*
0.000
0.000
16 4
1.000
1.000
1.000
1.000
1.000
1.000
*
*
*
1.000
1.000
1.000
1.000
1.000
1.000
0
0
0
0
0
0
*
*
*
15 4
*
*
*
*
*
*
*
*
*
*

*
*
0
0
0
0
0
0
*
*
*
*=No characters are parsimony informative within this group because all characters are constant, all derived states are autapomorphic, or all derived states are in polymorphisms
"Node numbers refer to monophyletic groups based on the strict consensus tree presented in Figure 312
Character and tree statistics are based on parsimony informative characters only


201


361
1 hypothesize that the CALEOS model is the best choice (or would be if an
algorithm for doing an analysis with this model were available) for analyzing character
state polymorphisms where the basis for the polymorphism is genetic differences among
individuals of a taxon. Scenario 4 (Figure 340) illustrates a situation where a notable
difference arises between the CALEOS model and distinct state model. Consider a
situation where three character states exist (0,1 and 2) and at least one taxon includes
individuals that exhibit state 0, state 1, and intermediates. The question arises as to how
many steps should be required to go from 0&1 to state 2, if no assumptions of character
order are made (no assumption is made as to whether state 2 is closer to state 1 or to state
0). If the taxon exhibiting 0&1 has a fixed allele (or set of alleles) which does not vary in
that taxon but produces variability in phenotype in different individuals, 0&1 to any other
state should require one step (the distinct state model), since 0&1 represents a unique
character state based on a unique allele that produces the range of variation in the
polymorphism. However, if the basis for the 0&1 polymorphism is genetic
polymorphism, I would argue that 0&1 to 2 should require two steps while 0&1 to 1 or
0&1 to 0 should require one. If state 2 is closest to state 0, then more genetic changes
would be required to go from 0&1 to 2 than from 0 to 2, and likewise if state 2 is closest
to state 1, more genetic changes would be required to go from 0&1 to 2 than from 1 to 2.
Consequently, even if no assumptions of the order of evolution are made (states 0,1, and
2 can change to any other state in one step) a change from 0&1 to 2 would always require
two steps. A similar situation but with two polymorphisms is presented in scenario 6
(Figure 342) where the CALEOS model would suggest one of two possible tree
topologies is shorter while the distinct state model would not.


171
Figures 146-149. Male genitalia capsules of Chlosyne sp. In ventral aspect. 146: C. Ieanira.
147: C. Ieanira. 148: C. theona. 149: C. lacinia.
23(0/2 ntermediate)
23(0)
146
147


360
accounted for by an allele (or group of alleles) that produces different phenotypes in
response to different environmental stimuli (temperature and photoperiod are potential
examples). Examples of this in butterflies are almost certainly differences in phenotype
between different generations during the same year. For example, the spring brood of
Pier is napi in Wisconsin contains dark scaling on the ventral wing surfaces along the
veins while the summer broods do not (or at most very sparsely), first generation
individuals of Polygonia interrogationis in Wisconsin have one of two ventral wing
forms while second generation individuals have one of five different forms (and only
very rarely individuals with one of the first generation forms are collected), and during
the summer in Wisconsin and Florida almost all individuals of Junonia coenia are tan on
the ventral surface while during the cooler months many individuals are found with red
ventral wing surfaces and intermediate between tan and red (pers. obsv.). In an informal
experiment, when I had pupae of Junonia coenia reared from larvae collected during the
summer in Florida that I refrigerated for several days, some of the adults that emerged
were the red form, whereas pupae I did not refrigerate produced only tan individuals. It
seems highly unlikely that these and other phenotypic polymorphisms associated with
different generations during the same year or different times of the year are based on
genetic differences that consistently alternate between generations. While none of the
character state polymorphisms I coded within the Melitaeini are to my knowledge
associated with different generations or times of the year, it is possible some of these
polymorphisms are accounted for by fixed alleles that can produce a particular range of
variation as opposed to genetic variation among individuals of the same taxon.


363
Aside from the issue of whether characters with polymorphisms are more
homoplastic, Wiens (1999) argues that more characters are preferable to fewer even if the
characters added increase the proportion of homoplasy in the data set. This argument is
consistent with my comparison of congruency between trees derived from genitalic and
pattern characters. Individual pattern characters are clearly on the average more
homoplastic than individual genitalic characters (Figure332), yet there is clearly no
evidence to suggest that an analysis based on all pattern characters combined is less
accurate than one based on all genitalic characters combined or than one based on all
characters combined (see the section: Congruency between tree topologies derived from
genitalic and pattern characters).
I developed the DPCWH model for comparative purposes because it is feasible to
run a DPCWH analysis and not a CALEOS analysis in PAUP. The DPCWH model
guarantees to find all the shortest CALEOS trees (and consequently all the shortest
distinct state model trees as well); however, in some scenarios it considers one of two or
more alternative tree topologies to be shorter when they are considered the same length
by the CALEOS model. Also, the DPCWH model poses problems for character
weighting, either all single state taxa have changes for a polymorphic character weighted
twice, or all multistate taxa have changes weighted half. I elected to go with the latter
option since in general there are few multistate taxa relative to single state taxa for
polymorphic characters in my data set. The DPCWH model is also problematic with
respect to successive weighting, although I suppose after PAUP reweights the characters
one could change each character in a doublet to half of the weight assigned by PAUP (I
did not do this). Furthermore, this model could be problematic if there were multiple


Table 12 Continued
736
NMNH=Nat. Mus. of Natural Hist.-Smithsoman Inst.; FSCA-FL St. Collection of Arthropods; AM-Allyn Mus.; PC=Personal Collection; @=Judged by author as of dubious authenticity
County (USA)
Locality (USA)
State/Prov.
Locality (non US)
Locality Continued (non USA)
Dale
Chiosyne acaslus
USA Arizona 1 Pinal
Peppersauce Canyon
15 Mar 1980
Abner A Towers
FSCA
j
Chlosyne acaslus
USA Arizona Yavapai
Mescal Canyon, 2 mi SW of Jerome on Hwy 89A
25 April 1965
Kilian Roever
FSCA
Chios yne acaslus
USA Arizona
Yavapai
Water tanks above Jerome, 5600'
11 April 1963
FSCA
Chlosyne acaslus
USA California
Los Angeles
Mojave Desert, Bob's Gap
10 April 1963
Henry Chavez
NMNH
t
Chios yne acaslus
USA California
Los Angeles
Mojave Desert, Palmdale Area
2 April 1966
F P Sala
NMNH
i
Chlosyne acaslus
USA California
Riverside
Chino Canyon nr. Palm Springs
29 March 1960
T C Emmal
FSCA
4
Chlosyne acaslus
USA California San Bernad.no
Joshua Tree Nat. Monument
12 April 1958
Tom Emmel
FSCA
Chlosyne acaslus
USA California San Bernad.no
Rock Corral
27 April 1978
B Weber
FSCA
Chlosyne acaslus
USA California
Mojave Desert near Willow Springs
10 May 1958
T C Emmel
FSCA
3
Chlosyne acaslus
USA California
Mojave Desert
27 Mar 32
FSCA
Chlosyne acaslus
USA California
Mojave Desert
10 April 1932
L J Kerr
FSCA
Chlosyne acaslus
USA California
Mojave Desert
8 Apr 28
FSCA
1
Chlosyne acaslus
USA Colorado Delia
Black Ridge. 6500'
2 June 1965
FSCA
Chlosyne acaslus
USA Colorado Della
1 Mi N of Payne Sidings
15 May 1965
FSCA
Chlosyne acaslus
USA Colorado Mesa
10 miles E of Gateway Unaweco Canyon, 6200'
23 Aug. 1966
FSCA
Chlosyne acaslus
USA Colorado
Mesa Verde Pak
3 June 1938
FSCA
1
Chlosyne acaslus |Dk Femmes)
USA Oregon Wasco
2 miles S of Tygh Valley. 1100'
24 June 1967
FSCA
4
Chlosyne acaslus
USA Utah Beaver
Beaver Cr Hills
NMNH
Chlosyne acaslus
USA Utah Cache
Logan Canyon near Card Camfigrd., Wasatch Mis.
8 June 1963
Ken Tidwell
FSCA
Chlosyne acaslus
USA Utah Salt Lake
Mill Creek Canyon, Wasatch Mis., 5330"
23 May 1962
FSCA
2
Chlosyne acaslus
USA Utah Tooele
North Willow Creek. Slansbury Mis 5800'
18 May 1962
FSCA
S
Chlosyne acaslus
USA Utah Tooele
North Willow Creek. Slansbury Mis., 5600'
19 May 1963
Ken Tidwell
FSCA
Chlosyne acaslus
USA Utah Tooele
North Willow Creek, Slansbury Mis 6200'
31 May 1964
FSCA
1
Chlosyne acaslus
USA Utah Tooele
South Willow Crook. Slansbury Mis Low' Narrows Campground. 6800'
FSCA
Chlosyne acaslus
USA Utah Uintah
Merlleys Park
31 May 1963
FSCA
Chlosyne definita (FSCA.AM.NMNH(S of US border only))
Chlosyne definita
USA
Texas Cameron
Brownsville
NMNH
Chlosyne definila
USA
Texas Culberson
Kent
5.1V 1946
B Weber Codec lion
AM
5
Chlosyne definila
USA
Texas Nueces
Corpus Christi
March 16-23
Barnes
NMNH
Chlosyne definila
USA
Texas Nueces
Corpus Christi
May
AM
Chlosyne definila
USA
Texas Nueces
Corpus Christi
AM
Chlosyne definila
USA
Texas
Nueces
Nueces
AM
2
Chlosyne definila
USA
Texas
Pecos?
12 Mi N of Marathon. Glass Mins 5,000', Composit
12-IX-1960
Lee D Miller
AM
1
Chlosyne definila
USA
Texas
Presidio
2.7 miles S of Shatter
16 July 1968
John Hafernik
FSCA
Chlosyne definila
USA
Texas
Presidio
2 7 miles S of Shatter
18 July 1968
John Hafernik
FSCA
Chlosyne definila
USA
Texas
Presidio
2 miles S of Shatter
21 Aug. 1968
John Hafernik
FSCA
Chlosyne definila
USA
Texas
Presidio
2 miles S of Shatter
24 Aug. 1968
John Hafernik
FSCA
2
Chlosyne definita
USA
Texas
Presidio
2 miles S of Shaflen
20 June 1972
John Hafernik
FSCA
Chlosyne definila
USA
Texas
Reeves
Redd Bluff LKR+285
30-VII-1973
E C Olson
FSCA
Chlosyne definila
USA Texas
San Patricio
Orange Grove
IV-2-61
NMNH
20
Chlosyne definita
USA Texas
ChinoO Mountains
30.IX 1928
E R Tinkham
AM
Chlosyne definita
USA Texas |?
Franklin Mis.. McKelligan Park
2 April 1969
Jim Scott
FSCA
Chlosyne definila
Mexico 80 Mi NW of Mexico City, 8000
21 Aug 1971
R K Robbins
NMNH
Chlosyne definila
Mexico Campeche
Lerma
VII.54
T Escalante
AM
1
Chlosyne definita
Mexico Chihuahua
Santa Rosa
May'06
Schaus
NMNH
Chlosyne definila
Mexico Federal District
Atizapan
VIII 65
T Escalante
AM
Chlosyne definita
Mexico Hidalgo
Hwy Nr Jacala. 5,20
O'
9 March 1959
FSCA
Chlosyne definila
Mexico
Hidalgo
Jacala, 4500'
VI.24.39
R Haag
AM
Chlosyne definila
Mexico
Nuevo Leon
20 Mi N Monterrey
19 Aug 1979
I L Finkelstern
FSCA
,
Chlosyne definila
Mexico
San Luis Potos Ciudad Viles
19-X-1970
H L King
FSCA
1
Chlosyne definila
Mexico
Tamaulipas
C. Victoria
VI-19.39
AM
Chlosyne definila
Mexico
Veracruz
Region de Cordoba
1901-3
L Conradt
NMNH
Chlosyne definila
Mexico
Veracruz
Tierra Blanca
VIII.-29
T Escalante
AM
1
Chlosyne definila
Mexico
Veracruz
Tierra Blanca
VII-39
T Escalante
AM
Chlosyne definila
Mexico Veracruz
Tierra Blanca
VII-40
T Escalante
AM
Chlosyne definila
Mexico Coahuila OR Guerrero; San Marcos
July-07
W M Schaus
NMNH
Chlosyne definila
Mexico ?
1897
J N Rose
NMNH
Chlosyne definita
@Argentina @Entre Rios sato Grande
7.V.76
D W Jenkins
AM
Chlosyne leanira leanira (NMNH.FSCA)
Chlosyne leanira leanira
USA Arizona Coconino
@So. Rim Grand Canyon
1 June 1958
GW Rawson
NMNH
Chlosyne leanira leanira
USA California Almeda
Jun 16-23
NMNH
10
Chlosyne leanira leanira
USA California Inyo InyoMls
2? May 1937
NMNH
Chlosyne leanira leanira
USA California Inyo Olaicha
April 24-30
NMNH
Chlosyne leanira leanira
USA
California Los Angeles Mini Canyon
April 10 1920
NMNH
Chlosyne leanira leanira
USA
California Los Angeles Mini Canyon
May 12. 1920
J. Comstock
NMNH
3
Chlosyne leanira leanira
USA
California Los Angeles Mini Canyon
April 30 1923
NMNH
Chlosyne leanira leanira
USA
California Los Angeles Mini Canyon
May 5 1925
J. A Comstock
NMNH
Chlosyne leanira leanira
USA California Los Angeles Mini Canyon
May 8 1925
NMNH
Chlosyne leanira leanira
USA California Los Angeles Mini Canyon
May 9 1925
NMNH
Chlosyne leanira leanira
USA California Los Angeles Mini Canyon
14 April 1933
FSCA
Chlosyne leanira leanira
USA California Los Angeles ¡Mini Canyon
May 4 1933
NMNH
Chlosyne leanira leanira
USA California Los Angeles 'Mini Canyon
1 June 1933
Chlosyne leanira leanira
USA California Los Angeles Mini Cr
May 4 1931
P E D. Wright
NMNH
Chlosyne leanira leanira
USA California Los Angeles Pasadena
April 4 1901
NMNH
Chlosyne leanira leanira
USA California Los Angeles Pasadema
August 4. 1901
NMNH
Chlosyne leanira leanira
USA California
Loe Angeles
Los Angeles
San Fernando
April 1-7
Barnes
NMNH
Chlosyne leanira leanira
USA California
April 8-15
NMNH
Chlosyne leanira leanira
USA California Los Angeles
Coquillet
NMNH
Chlosyne leanira leanira
USA California Marin Fairfax Canyon
Jun61931
J E Cottle
NMNH
Chlosyne leanira leanira
USA California Mann Mill Valley
W A Hammer. J K Newlm
FSCA
Chlosyne leanira leanira
USA California Marin
Jun 16-23
Barnes
NMNH
Chlosyne leanira leanira
USA California Mariposa Incline
2 June 1962
FSCA
Chlosyne leanira leanira
USA California Maiposa Incline
II 1963
W A Hammer
Chlosyne leanira leanira
USA California Mariposa Indian Flat
30 May'60
TW Davies
FSCA
Chlosyne leanira leanira
USA California Mariposa Indian Flat, 1900
13 May 1961
Chlosyne leanira leanira
USA California Mariposa Indian Flat
May 13 1964
Ral Davis
NMNH
Chlosyne leanira leanira
USA California Mariposa Indian Flat
Davies
FSCA
Chlosyne leanira leanira
USA California Orange Orange County Park
2 June 1921
LTKerr
Chlosyne leanira leanira
USA California Orange Orange County Park
3 July 1921 L T Kerr
FSCA
Chlosyne leanira leanira
USA California Orange Orange County Park
1 June 1922 ¡LTKerr FSCA
Chlosyne leanira leanira
USA California Riverside Palm Springs
April 1-7 NMNH
Chlosyne leanira leanira
USA California San Benito Pinnacles
May 16, 1939
Chlosyne leanira leanira
USA California Sai Bernad.no Catn Pass
May 19 1927
NMNH
Chlosyne leanira leanira
USA California San Bernadino C*on Pass
May 1930
MMNH
Chlosyne leanira leanira
USA California j San Bernadino
3ajon Pass
0 April 1934
Chlosyne leanira leanira
USA California San Bernad.no
-amp Baldy. San Bern Mis
NMNH
Chlosyne leanira leanira
USA California Siai Bernadino
>o Grande
3 April 1933
Chlosyne leanira leanira
JSA
california Isan Bernadino
an Bernandino
un 11 1895
NMNH
1
Chlosyne leanira leanira
USA California San Bernadino
Chlosyne leanira leanira
USA California 'Stpi.
t


291
pattern elements are present, the postmedian band patches are in the same position
relative to each other. The cell CuA2 patch is located near the wing margin, and often
partially fused with the submarginal band (element "j" in Nijhout (1991)) (if present).
The relative placement of other patches is as follows: cell CuAl patch distinctly basal to
cell CuA2 patch, M3 patch distal to cell Cua2 patch at the lower comer or about even
with it, cell M2 patch distinctly distal to cell M3 patch, cell Ml patch distinctly basal to
cell M2 patch, cell R5 patch distinctly basal to cell Ml patch, cell Sc+Rl patch distinctly
basal to cell R5 patch.
The color of the hindwing postmedian band is always identical to that of the
following features (if present) in any given specimen: the inside of the discal spot, the
area between the distal band of the basal symmetry system and band of the proximal
symmetry system in the forewing discal cell, the inside of the basal discal cell patch on
the forewing, the forewing postmedian band, and the basal patch in cell Cu2. The
postmedian band is the most conserved of all of these features, and is always present in
any of the taxa which possess any one of them (except for Chlosyne lacinia).
Consequently, I code the color of the hindwing postmedian band and not of the other
above features, as their color relative to the postmedian band clearly seems to be
dependent.
The two basic colors found among taxa for the hindwing upperside postmedian
band are orange and cream colored. Among those taxa with orange there is variation in
the color of orange. Chlosyne endeis, C. melanarge, C. poecile, C. marina, C. eumecla,
and C. erodyle always have a distinctly red-orange upperside hindwing postmedian band.
Chlosyne harrissii, C. gorgone, C. nycteis, and C. definita always have a light orange


345
proclea (Doubleday)
pelops (Drury)
Microtia Bates
eleda (Hewitson)
elva Bates
dymas (Edwards)
coracara (Dyar)
anmalas (Godman & Salvin)
Cltlosyne Butler
harrisii (Scudder)
kendallorum Opler
nycteis (Doubleday)
gorgone (Hubner)
hoffmanni (Behr)
palla (Boisduval)
whitneyi (Skinner)
a. damoetas (Skinner)
b. whitneyi (Behr)
gabbii (Behr)
acastus (W. H. Edwards)
defmita (Aaron)
ezra (Hewitson)
chinatiensis (Tinkham)
perlula (C. & R. Felder)
theona (Menetries)
lean ira (Felder & Felder)
a. leanira
b.fulvia (Edwards)
c. cyneas (Godman & Salvin)
d. cynisca (Godman & Salvin)
endeis (Godman & Salvin)
a. endeis
b. pardelina Higgins
marina (Geyer)
melitaeoides (Felder & Felder)
erodyle (Bates)
a. erodyle
b. poecile (Felder & Felder)
melanarge (Bates)
eumeda (Godman & Salvin)
hylaeus (Godman & Salvin)
californica (Wright)
lacinia (Geyer)
ehrenbergi (Geyer)
hippodrome (Geyer)
narva (Fabricius)


104
E. chalcedona, E. colon, and E. anida, with E. editha as the sister taxon to this clade.
These characters include an anterior orientation of the dentate side of the dorsal fork of
the inner valve process, the orientation of the tip of this fork strongly ventral and weakly
inward, the orientation of the tip of the ventral fork of the inner valve process curved
ventrally outward, and the granulse patches on the tegumen very conspicuous at 50X
magnification. All of these characters represent terminal derived states of multistate
characters o the most parsimonious tree.
Scott (1986) suggested E. anida, E. colon, and E. chalcedona are all members of
one species which exhibits clinal variation in genitalia and wing pattern. While I did not
investigate this hypothesis, the phylogenetic analysis does support these three taxa as a
monophyletic group based on the terminal derived character states of the posterior juxta
projection being far more heavily sclerotized than the rest of the juxta, and the posterior
juxta projection not being laterally compressed (the latter is a homoplastic character).
Distribution: Euphydryiti is the only subtribe where representatives occur in both the Old
and New Worlds, including closely related taxa in the genus Hypodryas. Euphydryas is
exclusively new world. The eastern species (E. phaeton) apparently does not overlap
with any of the western taxa, and occurs in the northeast and midwest United States as far
south as southern Georgia and as far west as eastern Texas and Nebraska. Other
representatives occur throughout the western United states as far east as western South
Dakota and Nebraska, and in northwest Mexico (Stanford & Opler 1993) and southwest
Canada. Hypodryas is primarily an Old World genus, but a single representative occurs
in Idaho, Montana, Wyoming, and in southern Alberta and British Colombia (Stanford &


626
M3 and smaller orange patches in cells CuAl and M2. In addition, the dorsal forewing
anal cell usually has an elongate diffuse orange patch originating about even with the
basal edge of the cell CuA2 median band patch and extending basal beyond the basal
cream patch in cell CuA2. Thin orange scaling usually borders the periphery of the distal
discal spot cream patch, at least on the basal and distal edges. Many specimens have
little or no diffuse orange scaling between the postmedian and median bands or basal to
the median band, especially females, but a portion of specimens have more extensive
orange in this area. The most extreme examples have orange patches distal to the
postmedian band in cells CuA2 through Ml, with diffuse to sharp orange streaks in these
cells between the median and postmedian bands and sometimes extending basal beyond
the median band, in addition to some diffuse orange scaling in the discal cell. Even in
these oranger specimens, the area distal to the postmedian band in cells R5 through R.3 is
predominately black (sparse scattered orange scales may be detectable with magnification
only), and cream scales are not present distal to the postmedian band in any cell.
The dorsal hindwing has a cream colored median and postmedian band, and a
variable amount of orange. In males, the median band patch in cell CuA2 is often absent
or diffuse. The hindwing postmedian band may have little black along the veins between
the patches and appear as a band, or have the patches reduced to small dots such that they
resemble (in shape, not color) the white row of postmedian dots found in some more
derived Chlosyne. The hindwing median band is almost always narrower in basal to
distal width relative to subspecies fulvia, but oranger cyneas specimens also tend to have
a wider median band and those cyneas specimens with the broadest median band have the
median band width about equal to the fulvia specimens with the narrowest median band.


165
Character numbers are from the Chlosyniti data matrix (Table 5)
Figures 108-113. Male genitalia capsules of Microtia sp. 108: M. eleda (ventral). 109:
Same (dorsal). 110: M. elva (ventral). Ill \ M. elva (dorsal). 112: M. coracara (ventral)
113: M. coracara (dorsal)


784
PLATE Q


I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation
for the degree of Doctor of Philosophy.
Thomas C. Emmel, Chair
Professor of Entomology and Nematology
I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation
for the degree of Doctor of Philosophy.
John B. Heppner
Assistant Professor of Entomology and Nematology
I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation
for the degree of Doctor of Philosophy.
/James E. Lloyd
Professor of Entomology and Nematology
I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation
for the degree of Doctor of Philosophy.
Professor of Entomology and Nematology
I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is
for the degree of Doctor of Philosophy.
Jonathen Jteiskind
Associate Professor of Zoology
fully adequate, in scopgand quality, as a dissertation
This dissertation was submitted to the Graduate Faculty of the College of Agriculture and
to the Graduate School and was accepted as partial fulfillment of the requirements for the degree
of Doctor of Philosophy.
December, 2000
)ean, College of Agriculture
Dean, Graduate School


701
subspecies as noted below. The remainder of the dorsal forewing is black except for the
postmedian dots. The dorsal forewing postmedian dots are conspicuous and pure white,
present in cells CuAl-R5 and R3 but not CuA2. The postmedian dots appear to occur
farther basally than in most Chlosyne, perhaps due to the unusual wing shape, and C.
gaudealis has no forewing median band. Consequently, it was difficult to know for
certain a priori if the white forewing dots in C. gaudealis are homologous to postmedian
dots or median dots of other Chlosyne. The CuAl and M3 dots being far basal to the M2
dot is characteristic of the median band in some Chlosyne, but the postmedian dots of C.
narva (which also has a clear median band), the sister species to C. gaudealis, are in
similar in position to C. gaudealis. Furthermore taxa with the median band occurring as
a row of dots have the R5 median dot far basal to the Ml median dot, while the R5 dot is
only slightly basal to the Ml dot in C. gaudealis, characteristic of the postmedian dots in
C. narva and other Chlosyne. Consequently, I interpret the row of forewing dots in C.
gaudealis as postmedian dots. Ventrally the forewing markings are the same as dorsally
(except for below subspecies differences in basal red scaling), with few exceptions. A
series of yellow patches are present ventrally from cells Ml (short) through R3 (longer)
which occurs in the same position as the portion of the ventral median band occupying
these cells in C. narva (although the Ml section is larger in this taxon) and consequently
I interpret this feature as a remnant of a ventral median band. In a few specimens
examined of subspecies gaudealis a smaller patch of yellow occurs on the dorsal
forewing surface in the same position. Also, there are pale cream to yellowish cream
patches of variable number and distal to the postmedian dots. These patches tend to be


399
Table 1
0:
Variation in boot strap s
cores with different numbers of taxa
Nod* in
rw(yw
1
lili 1 1
No.
Node Boot Strap Score
s*
NC=
Not calculated because the group is outside the node in
the
anal^
is or in a basal clade
Node
Taxa
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21a
21b 22
23
24
25
26
27
26
29
30
31
32 33
34 35 36
37 38 39
40
41
4.
4 -
44
45
46
1
54
by
8
n
100
100
65
43
45
97
79
66
45
60
40
99
78
84
62
85
77
41
X
82
93
87
85
53
45
65
22
25
84
70 26
58 38 98
75 61 80
90
87
86
96
70
59
84
4
48
NC
NC
NC
NC
NC
60
X
X
98
NC
82
47
67
42
99
72
81
61
87
74
X
36
84
94
86
85
54
45
62
18
23
84
73 27
57 56 97
76 61 81
90
89
86
96
71
59
81
7
46
NC
NC
NC
NC
NC
NC
48
51
98
NC
NC
59
72
46
99
67
83
59
88
73
X
36
83
93
86
86
55
47
64
20
25
86
70 28
57 36 98
74 60 81
90
87
86
97
69
60
85
8
41
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
76
47
100
39
83
57
NC
96
46
X
83
94
88
84
53
46
67
18
26
87
71 30
<0 I 40 I 97
76 62 84
91
87
86
96
69
60
80
12
31
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
X
46
88
95
91
85
46
43
60
15
22
87
71 26
52 36 97
73 62 83
91
88
87
96
70
60
83
13
30
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
96
96
86
36
37
54
X
21
88
71 23
47 33 97
74 63 83
91
88
86
96
69
61
99
17
22
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC NC
NC
NC
NC
NC
NC
NC
NC NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
X
52
83
72 59
90 67 98
86 63 84
89
88
83
96
69
60
NC
18
20
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
57
84
65 54
94 i 71 I 97
84 63 83
87
87
84
95
69
60
NC
19
18
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
71 NC
97 67 97
82 61 83
86
88
83
95
69
59
NC
20
15
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC NC NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC NC
NC 85 97
92 59 86
91
89
82
93
68
60
NC
21
14
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC NC
NC'NC 96 92 X l 86
92
NC
81
90
68
60
NC
9
10
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
37
100
NC
91
63
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC NC
NC NC NC
NC NC NC
NC
NC
NC
NC
NC
NC
NC
14
8
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
89
NC
51
48
NC
NC
NC
NC
NC NC
NC NC NC
NC NC NC
NC
NC
NC
NC.
NC
NC
NC
22
7
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC NC
NC NC NC
NC 71 95
96
NC
NC
NC
NC
NC
NC
25
7
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC NC
NC NC NC
NC NC NC
NC
NC
NC
NC
72
65
NC
2
6
NC
100
97
100
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC NC
NC NC NC
NC NC NC
NC
NC
NC
NC
NC
NC
NC
11
6
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
63
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC NC
NC NC NC
NC NC NC
NC
NC
NC
NC
NC
NC
NC
3
5
NC
NC
94
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC NC
NC NC NC
NC NC NC
NC
NC
NC
NC
NC
NC
NC
5
5
NC
NC
NC
NC
NC
NC
NC
79
100
NC
NC
NC
NC
NC NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC NC
NC NC NC
NC NC NC
NC
NC
NC
NC
NC
NC
NC
23
5
NC
NC
NC NC
NC
NC
NC
NC
NC
NC
NC
NC
NC NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC NC
NC
NC
NC NC
NC NC NC
NC NC 97
NC
NC
26
5
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NCl NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC NC
NC NC NC
NC NC NC
NC
NC
NC
NC
NC
NC
6
4
NC
NC
NC
NC
NC
NC
NC
NC
100
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC NC
NC NC NC
NC NC NC
NC
NC
NC
NC
NC
10
4
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
100
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC NC
NC NC NC
NC NC NC
24
4
NC
NC
NC
NC
NC
NC
NC
NC
NC NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC; NC
NC NC NC
NC NC NC
99
NC
NC
NC
NC
NC
NC
n
1
2
3
2
1
2
2
3
5
1
2
3
4
i
6
4
5
6
3
4
2
3
5
6
7
6
7

6
5
B
8
9 8
9 10 11
11 11 13
14
'
11
12
12
6
High
69
100 97
100
100
65
48
79
100
79
82
59
76
47
100
78
91
63
88
96
46
46
88
96
96
86
55
48
67
22
57
88
73 59
97 85 98
92 71 97
99
89
87
97
72
70
99
Low
69
80
78
100
100
60
43
45
97
79
66
45
60
37
99
39
81
57
85
73
41
36
82
93
86
84
36
37
54
15
21
83
65 23
47 33 96
73 59 80
BG
87
81
90
68
59
80
Range
0
20
19
0
0
5
5
34
3
0
16
14
16
10
1
39
10
6
3
23
5
10
6
3
10
2
19
11
13
7
36
5
8 36
50 52 2
19 12 17
13
2
6
7
4
11
19
Mean
NA
00
697
100
NA
62 S
455
563
6 6
NA
74 0
SO 3
68 6
42 4
MS
640
644
606
66 7
600
NA
NA
640
42
so
652
44 4
444
2 0
186
31 4
854
70 4 34 1
6 0 52 9 97 2
80 4 62 4 85 1
Standard Dev
NA
M I
102
00
NA
}J
1.0
16 1
13
NA
11.3
76
6 6
42
05
17 3
36
2.4
IS
106
NA
NA
23
12
36
06
7
36
4
26
14 4
16
22 14 0
19 7 166 06
72 11 52
36
06
Median
69
io
94
100
100
63
46
51
98
79
74
47
70
42
100
70
83
62
87
76
44
36
83
94
88
85
53
45
63
18
25
85
71 28
58 48 97
76 62 83
91
88
86
96
69
60
84
in Figure 311

Node numbers refer to the clades designated in Figure 312


289
having a small diffuse light colored patch along the upper vein to having the wing
entirely black, including Chlosyne ezrci and Chlosyne lacinia. Consequently, the
variation in wing pattern in this section of the wing appears to be too variable
intraspecifically to provide reliable phylogenetic information which can be definitively
coded into discrete states.
97. The distal element of the basal symmetry system in cell CuA2 (dorsal and ventral
surfaces):
0=Double, forming a patch with light scales inside of identical color to the
forewing discal spot or solid black.
l=Single, forming a narrow, elongate patch with a black border extending from
the basal part of the cell almost to the origin of vein CuA2.
2=Absent.
Those taxa in the Thessalia group which had character state 2 for character 88 exhibit the
same type and range of variation for the CuA2 basal patch. Coding these taxa as a
separate state for this character would likely have weighted the same evolutionary change
twice. Since the underside reveals these taxa have the CuA2 patch as the derived form
(state 1) I code these taxa as state 1 based on the ventral surface only.
98. For those with the distal band of the basal symmetry system double and fused into a
patch, the occurrence of light colored scaling inside of this patch (of identical color to
that inside of the discal spot):
0=Present.
#l=Absent, the patch is solid black.
@A=0&1: Varies along a continuum between state 0 and state 1.


84
that obtained from equally weighted characters, except there is greater resolution within
the Phycioditi clade.
Boot strap 50% consensus trees from equally (Figure 295) and successively
weighted characters (Figure 296) with the representative out group method support the
same clades with one notable exception. The clade with Chlosyne comes out as sister
clade to the clade with Poladryas with a score of 51% for successively weighted
characters, but this clade only has a score of 40.5% for equally weighted characters.
Most clades from the two analyses have scores within 5% of each other, and only one
other clade has a score with a difference greater than 9%; the clade uniting the two
representatives of Cinclidia and Melitaeci has a score 13% lower for equally weighted
characters. None of the clades supported by the boot strap analyses are in conflict with
the strict consensus trees for equally or successively weighted characters, and all of the
clades appearing on the strict consensus tree of equally weighted characters appear on
both 50% boot strap consensus trees with three exceptions. One exception is the missing
union of the Chlosyne group and Poladryas group on the boot strap tree from equally
weighted characters as noted. In addition, the relationships between the three clades
coming out as the sister clade to Euphydryiti are unresolved on both boot strap trees.
Finally, Mellicta varia, M. asteria, and M. aurelia came out as a resolved clade on the
strict consensus trees but this clade was collapsed on both of the boot strap consensus
trees.
With the cumulative out group method, the heuristic search of equally weighted
characters yielded 2,546 equally parsimonious trees with a consistency index of 0.863,
retention index of 0.963, and rescaled consistency index of 0.831. The strict consensus


454
29. Orange hairs and scales present on the labial palpi, ventral forewing discal cell
predominately orange, area between the proximal and distal elements of the central
symmetry system on the ventral hindwing containing orange at least in the discal cell....
30
29. Orange hairs and scales absent on the labial palpi, ventral forewing discal cell with
no or sparse (some C. marina) orange, area between the proximal and distal elements of
the central symmetry system on the ventral hindwing lacking orange in all cells 32
30. Vertex and abdomen with orange scales, outer lateral side of labial palpi with a white
band and orange hairs and scales (few if any black scales/hairs) Chlosyne defmita
30. Vertex and abdomen without orange scales, outer lateral side of labial palpi with a
white band and orange and black scales Chlosyne endeis 31
31. Forewing discal cell with pale yellow scaling and often orange within both the discal
spot and the patch formed between the proximal element of the central symmetry system
and distal element of the basal symmetry system, ventral forewing with extensive orange
scaling Chlosyne endeis pardelina
31. Forewing discal cell black and at most with 2-3 small patches of cream colored
scales, ventral forewing with orange scaling limited to primarily the discal cell and
sometimes around the postmedian dots in cells CuA2 and Ml... Chlosyne endeis endeis
32. Dorsal hindwing solid black except for postmedian band patches, dorsal forewing
median band confined to R cells and cells M2 and Ml and dorsal forewing postmedian
band confined to cells Cua2-M3 (a minute patch sometimes present in M2), creating the
appearance of a single yellow band bordering the dorsal forewing apex from the outer
anterior comer to the costa distal to the discal cell Chlosyne melanarge


297
some specimens seem somewhat intermediate between state 0 and state 2, while others
appear to be either state 0 or state 2. I coded these taxa as 0&2. Also, the ventral
hindwing postmedian band was difficult to interpret in the out group taxa of genus
Higginsius, and I coded both of these taxa as
Note that references to the background in parenthesis are given for information
purposes only, they are part of the state delimitations.
104. For those taxa with state 2 of the preceding character, the color of the light patches
inside of the dark band relative to the background color of the ventral hindwing:
l=The light colored patches are the same color as the background.
2=The light colored patches are orange in contrast to the background color.
Taxa lacking character state 2 for the preceding character are coded ? (there is no
evidence to suggest state 2 of the preceding character is a terminal derived state).
All taxa examined with character state 0 of the second preceding character have a
secondary patch of color inside the series of light colored patches forming the ventral
hindwing postmedian band. The size of the secondary colored patches varies greatly
intraspecifically; however, the color of these secondary patches is uniform within a taxon
with one exception, Chlosyne nycteis.
105. For those taxa with state 0 of the second preceding character, the color inside the
ventral hindwing postmedian band patches:
0=Orange.
l=Brown.
@*A=0&1: Individuals are either state 0 or state 1.


268
Regarding state 2, the number of white scales and hairs was found to be small relative to
the number of black ones, but given this the extent of white seemed to vary
intraspecifically to very few or even none. It is possible some taxa may never have white
present, and some may vary along a continuum from having sparse white scales and hairs
to none. However, this is essentially impossible to determine as some scales are rubbed
off in this area in virtually every specimen, as this is the area where specimens are
pinched when collected in the field. Thus, coding separate states for specimens having
white scales in this area versus none could not be done accurately. The safest coding
appears to be including those taxa with primarily black scales and hairs in these grooves
(and no orange) in one state, rather than trying to split this state up by difficult or
impossible to interpret criteria. Chlosyne kendallorum is coded "2?" because this
character can only be examined in the dorsal half of the thorax of one specimen-one
specimen is denuded and the other is covered with glue over the ventral half of the
pleurites. Some specimens of Microtia elva may have few orange scales, and some even
appeared to have none but they were partly denuded in this area.
79. Color of scales and hairs on lateral thorax ventral to the wings and excluding the
groove where the femur tucks in.
0=Densely covered with white hairs and scales, some dark scales under the white
hairs.
l=White and pale orange hairs, some black scales underneath these hairs.
2=Densely covered with white and black hairs and scales with dark hairs more
concentrated dorsally, and some orange hairs and scales in the vicinity of the
coxae.


572
perlula, the cream background color distal to the postmedian band is a distinctly different
color (bright whitish cream) than the background color basal to the postmedian band
(darker yellowish or orange tinged cream) while in C. chinatiensis both areas are bright
whitish cream. The dorsal hindwing is similar in both taxa, although three out of four
specimens of C. pedula have the median band bordered by black distally which is never
the case in C. chinatiensis. Also, the band of black basal to the ventral hindwing
postmedian band (fused together basal parafocal elements) is always thin in C.
chinatiensis, while in C. pedida it is much broader, and nearly as wide or wider than
sections of the postmedian band in some cells.
C. chinatiensis can not be separated from C. theona, C. pedula, or C. ezra based
on genitalic characters.
Further Description: Labial palpi as in C. ezra (above). Vertex with a distinct white
centered patch reaching the posterior edge of the plate, black with scattered orange scales
elsewhere. Sutures at the lateral edges of the vertex with black and white scales. Tuft of
white hairs on the frontoclypeus anterior-lateral of each antennal base well developed.
Frontoclypeus with a variable pattern composed of black, white, and orange scales and
hairs with the white scales sometimes coalesced to a variable degree into a white stripe
below and centered between the antennal bases. Patch of orange scales present at the
edge of the inner anterior-lateral margin of the eye. Black and white but no orange scales
present in the collar between the head and pronotum. Posterior edges of posterior
abdominal segments with bands of white scales on dorsal and lateral sides, black and
scattered orange scales anterior to these bands. Ventrally abdomen with two longitudinal
parallel black stripes against white to yellowish cream background, with the midline


679
Arizona, which is surprising since culjutrix occurs in California, New Mexico, and Texas.
In California I have seen numbers of both forms adjutrix and crocale as well as numerous
intermediates, plus the one specimen of form lacinia noted above. Nevada and Utah
specimens include form crocale and crocale/adjutrix intermediates. I have seen only
form adjutrix from Oklahoma.
All of the forms occur in Mexico except for saundersi and its melanic form
paupera, although I have seen two Mexican specimens (from Veracruz and San Luis
Potosi) that are form lacinia but slightly intermediate toward form saundersi. I have seen
adjutrix from the states of Chihuahua, Sinaloa, Jalisco, Nuevo Leon, San Luis Potosi,
Tamaulipas, Vera Cruz, Hidalgo, Oaxaca, and Yucatan indicating a very widespread
distribution extending deep into southern Mexico. The approximate southern and
western extent of the area where this form occurs appears to be Oaxaca and Yucatan,
respectively. I have seen quehtala/adjutrix integrades only from the states of Veracruz
and San Luis Potosi, although only one specimen closest to quehtala from the latter. I
have seen few Mexican adjutrix/crocale integrades, and only from San Luis Potosi. I
have also seen few examples of Mexican crocale, and states where this form and
crocale/quehtala integrades occur include Sinaloa, Colima, Guerrero, and Jalisco, all in
western Mexico. I have seen form quehtala from these states in addition to Veracruz,
Morelos, and Jalisco, suggesting a widespread distribution of this form in southern
Mexico. However, it is apparently absent in northern Mexico, with Venadio, Sinaloa
being the most northerly record 1 have seen. I have seen form lacinia from the states of
Veracruz, San Luis Potosi, Chiapas, Morelos and Oaxaca. The only Mexican State where
I have seen lacinia/quehtala integrades in Veracruz, with lacinia occurring primarily


566
the sides; inner lateral side with white scales and hairs (some black may be mixed in)
dorsally and at base and black scales and hairs ventrally and at tip. Vertex with a distinct
white centered patch reaching the posterior edge of the plate, only black scales elsewhere.
Sutures at the lateral edges of the vertex with black and white scales. Tuft of white hairs
on the frontoclypeus anterior-lateral of each antennal base well developed.
Frontoclypeus with white patches touching the eyes ventro lateral of the antennal bases
and a white vertical stripe centered on the face below the midpoint of the area between
the antenna and black scales and hairs elsewhere. Patch of orange scales present or
absent at the edge of the inner anterior-lateral margin of the eye. Black and white but no
orange scales present in the collar between the head and pronotum. Posterior edges of
posterior abdominal segments with bands of white scales on dorsal and lateral sides,
black scales anterior to these bands. Ventrally abdomen with two longitudinal parallel
black stripes against a pale cream background, usually not appearing broken between
each segment. Tibia and tarsi of prothoracic legs with orange scales dorsally and white
scales and hairs ventrally. Femur of meso- and metathoracic legs with predominately
orange scales dorsally and white scales ventrally except at distal end where only orange
scales are present, tibia and tarsi with orange scales only. Meso- and metapleurons where
the femur tucks in with black scales and hairs, with white scales sometimes visible
underneath the dark ones. Thorax ventral to the wings with dense white and black hairs
with the black hairs more concentrated dorsally. Antennal shaft with many areas bare of
scales, elsewhere a checkered pattern of black and white scales with or without sparse
orange scales. Club with a white patch on the outer lateral side when scales present in
this area, but usually this area is bare.


271
@A=0&1: Individuals may be either state 0 or state 1, with variation in the
number of orange scales present between state 1 individuals.
Chlosyneperlula is coded "?" for the above character. While the two specimens
examined for this character lacked orange scales, its closest relatives (confirmed a
postiori) have some infividuals with and some individuals without orange scales on the
antennal shaft, thus from the limited sample it can not be determined if C. perlula should
be coded "0" or "0&1".
81. For specimens with scales present on the antennal club, a white patch on the outer
lateral side of the club:
0=Present.
*l=Absent, only black scales present.
The three Atlantea tulita specimens available had the antennae too denuded of scales to
score this character.
Wing Pattern Characters
Within the Chlosyniti and Poladryiti there is great variability and complexity in
the wing pattern on both the dorsal and ventral surfaces. While at first some wing
patterns between different taxa may seem quite different, a careful examination of what
position various wing patterns occur in relative to the venation and cells allows one to
code much of this variation into multistate characters for wing pattern elements which are
very likely to be homologous derivations of the various elements of the Nymphalid
ground plan figured in Nijhout (1991). A study of the wing patterns across the taxa has
revealed transitions between a number of primitive wing pattern elements and
progressively more derived forms. Wing pattern elements supply by far the greatest


247
2=The sides are flared out as very prominent posterior/lateral extensions, and the
ventral midpoint of the ridge is barely produced distal to the rest of the lamella
antevaginallis (Figures 279 & 281).
Independent evidence from other characters suggests state 1 of the preceding character is
a terminal derived state. Taxa lacking state 1 of the preceding character are coded 0.
49. The formation of a pouch by the lamella postavaginallis and lamella antevaginallis
around the opening to the ductus bursae:
#0=Not forming a pouch, projecting in opposite or widely divergent directions
(Figures 274, & 277-278).
l=Forming a partial open pouch, with the separation wide enough that the
opening to the ductus bursae can be seen in ventral posterior view without prying
the plates apart with a forceps, projecting at an acute angle (Figures 279-282 (on
the average pouch more open) & Figures 283-292 (on the average the pouch more
closed).
Upon casual inspection, Atlantea may appear to have a closed pouch. However,
the pouch in Atlantea is formed by a posteriorly curved broad emarginate process
extending off of the anterior edge of the lamella antevaginallis and fused with the lamella
postvaginallis at its base (Figure 275). The lamella antevaginallis itself projects opposite
the lamella postvaginallis typical of state 0. In taxa where the lamella antevaginallis plate
is angled acutely with respect to the lamella postvaginallis, the abdominal stemite is
fused to the distal end of the pouch. In Atlantea, where the pouch is formed by an extra
process, the abdominal stemite is attached to the base of the lamella antevaginallis plate,
and the process is free of the stemite throughout its length.


Table 1--Continued
137
Characters
Taxa
81
82
83 84 85
86 87
88 89 90
91
92 93 94 95
96 97 98 99
100
Chlosynel
1
2
1
1
0
1
2
1
3
1
1
0
2
1
1
2 0 0
1
1
Chlosyne2
1
2
1
1
0
1
2
1
3
1
1
0
2
1
1
2 0 0
1
1
Texola elada
1
2
1
1
0
1
2
2
3
0
1
0
2
1
1
? 1 1
1
1
*Microtia elva
1
2
1
1
0
1
2
2
T]
0
1
0
2
1
1
1 1 1
1
1
Dymasia dymas
1
2
1
1
0
1
1
?
T|
?
1
0
2
1
1
1 1 2
1
1
Texola corocara
1
2
1
1
0
1
?
?
?
1
0
2
1
1
1 1 2
1
1
Texola anomalus
Antillea pelops
1
2
1
1
0
1
1
?
?
?
1
0
2
1
1
1 0 0
1
0
Antillea proclea
1
2
1
1
0
1
1 ?
?
?
1
0
2
1
1
1 0 0
1
0
*Polydryas minuta
1
2
1
1
0
1
2
1
3
0
1
0
2
1
1
1 0 0
1
0
*Higginsius fasciatus
Higginsius miriam
1
2
1
2
2
?
2
2
5
0
1
0
2
1
1
1 0 0
1
0
Atlantea tulita
1
2
1
2
4
?
2
2
5
0
1
0
2
1
1
1 0 0
1
?
Atlantea pantoni
1
2
1
2
4
?
2
2
5
0
1
0
2
1
1
1 0 0
1
?
Mellicta athalia
1
2
1
1
0
1
2
1
2
0
1
1
2
1
1
1 0 0
1
2
Mellicta britomartis
1
2
1
1
0
1
2
1
T|
0
1
1
2
1
1
T|o o
1
2
Mellicta aurelia
1
2
1
1
0
1
2
1
2
0
1
1
2
1
1
1 0 0
1
2
Mellicta varia
1
2
1
1
0
1
2
1
Tt
0
1
1
2
1
1
1 0 0
1
2
Mellicta parthenoides
1
2
1
1
0
1
2
1
TT
0
1
1
2
1
1
10 0
1
2
Mellicta asteria
1
2
1
1
0
1
2
1
3
0
1
1
2
1
1
1 0 0
1
2
Mellicta alatauica
1
2
1
1
0
1
2
1
2
0
1
1
2
1
1
1 0 0
1
2
Cinclidia phoebe
1
2
1
1
0
2
2
1
2
0
1
1
2
1
1
10 0
1
2
Cinclidia scotosia
1
T
1
1
0
2
2
1
2
0
1
1
2
1
1
1 0 0
1
2
*Melitaea cinxia
1
2
1
1
0
2
2
1
2
0
1
1
2
1
1
1 0 0
1
2
Melitaea diamina
1
2
1
1
0
2
2
1
2
0
1
1
2
1
1
1 0 0
1
2
*Didymaeformia didyma
1
2
1
1
0
2
2
1
2
0
1
1
2
1
1
1 0 0
1
2
Didymaeformia trivia
1
2
1
?
0
2
2
1
2
0
1
1
2
1
1
1 0 0
1
2
Phyciodes tharos
2
2
1
2
3
?
2
2
3
0
2
0
2
2
2
1 0 0
2
0
Phyciodes batesi
2
2
1
2
3
?
2
2
TT
0
2
0
2
T"
2
1 0 0
2
0
Phyciodes phaon
2
2
1
2
3
?
2
2
TT
0
1
0
2
2
2
1 oTo
2
0
*Phystis simois
2
2
1
2
3
?
2
T
TT
0
1
0
2
T
2
1 0 0
2
0
Anthanassa texana
2
2
1
2
3
?
2
2
TT
0
1
0
2
2
2
1 0 0
2
0
Anthanassa tulcis
2
2
'T
2
3
?
2
T
TT
0
1
0
2
2
2
1 0 0
2
0
*Telenassa teletusa
2
2
2
2
3
?
2
2
Tt
0
1
0
2
2
2
1 0 0
2
0
Castilia castilla
2
2
2
2
3
?
2
1
TT
0
1
0
2
2
2
IT o o
2
0
*Dagon catula
2
2
2
2
3
?
2
2
3
0
1
0
2
2
2
T|o o
2
0
*Ortilia liriope
2
2
T
2
3
?
2
1
TT
0
1
0
2
2
2
"TTo o
2
0
Ortilia ithra
?
2
1
2
3
?
2
1
TT
0
1
0
2
2
2
1 0 0
2
0
Eresia eunice
2
2
2
2
3
?
2
1
3
0
1
0
2
2
2
1 0 0
2
0
Eresia frisia
2
2
1
2
3
?
2
2
3
0
1
0
2
2
2
1 0 0
2
0
Janatella leucodesma
2
2
2
2
3
?
2
1
3
0
1
0
2
2
2
T[o o
2
0
Janatella fellula
2
2
2
2
3
?
2
1
3
0
1
0
2
2
2
1 0 0
2
0
Mazia amaznica
2
2
1
2
3
?
2
1
3
0
1
0
2
2
2
1 0 0
2
0
Tegosa claudina
2
2
1
2
3
?
2
2
3
0
1
0
2
2
2
1 0 0
2
0
Tisona saladellensis
2
2
1
2
3
?
2
2
3
0
1
0
2
2
2
1 0 0
2
0
*Gnathotriche exclamationis
1
2
1
2
1
?
2
2
4
0
1
0
2
1
1
1 0 0
2
0
Gnathotriche sodialis
TT-
*Gnathotrusia mundina


!
99
setae. In lateral view valvae orientated primarily anterior to posterior, widely separated
in ventral view. Surface ofjuxta smooth or with a mid ventral ridge, usually terminating
in a posterior process (composed of a anterior and posterior component close together
and appearing as a single process in most aspects). The single taxon examined lacking
this process (Eurodryas desfontaini) appears to have lost it due to reversal as opposed to
being the most primitive representative. Saccus forked with a pair of invaginated
projections. Tegumen narrow, not produced into a well developed uncus (although
paired posterior projections are present). Sclerotization pattern on the ventral surface of
the phallus with the posterior end lightly sclerotized, anteriorly more darkly sclerotized
along the sides with this area eventually widening into a concave border anteriorly and
originating closer to the supersensory membrane than the posterior phallus opening.
Dorsally the middle of the phallus is composed of a posteriorly tapering sclerotized
structure which extends as a triangular extension beyond the ventral edge of the posterior
phallus opening. Posterior edge of phallus lacking plates or attachments connected to the
vesica. Vesica simple other than for the presence of granulse patches, except for the
two Eurodryas examined which have a pair of disjunct sclerotized teeth bearing plates on
the ventral side of the everted vesica.
Females: Ventral plate on corpus bursae bearing teeth on the inner side. Well over half
of the length of the corpus bursae free of the ventral plate. Inverted teeth on the corpus
bursae not distinctly sclerotized, concolorous with the corpus bursae. Ductus bursae
sclerotized up to the junction with the ventral anterior genital opening. Ostium bursae
absent. Lamellae not enclosing the genital opening in a pouch, with the lamella
postvaginallis and lamella antevaginallis well developed and projecting in opposite


681
west that 1 have seen form sciundersi (excluding the perhaps questionable Honduras
record noted above).
In South America all specimens I have examined are form saundersi, form
paupera, or intermediates. I suspect paupera occurs wherever saundersi does, and while
I have not seen examples from all saundersi localities, the form appears to be
considerably less common in collections. I have seen saundersi from Colombia,
Venezuela, Peru, Ecuador, Trinidad, Brazil, Paraguay, Bolivia, and Argentina, and
paupera from Colombia, Venezuela, Ecuador and Bolivia.
Further Description: Black but no orange scales and hairs on the labial palpi. Dorsally
palpi with black scales and hairs only; outer lateral side white ventrally-center and at base
while black dorsally and at tip, black hairs around the edge with some white ones mixed
in on the ventral side; ventrally with base white and black hairs and scales in the center
and throughout the terminal segment, with the white lateral bands extending down along
the sides; inner lateral side with white scales and hairs (some black may be mixed in)
dorsally and at base and black scales and hairs ventrally and at tip (some specimens may
have sparse white scales mixed in at the tip). Vertex with a distinct white centered patch
reaching the posterior edge of the plate, only black scales elsewhere. Sutures at the
lateral edges of the vertex with black scales only. Tuft of white hairs on the
frontoclypeus anterior-lateral of each antennal base well developed. Frontoclypeus with
black scales and hairs and a highly variable pattern. Some individuals have a white
vertical stripe centered on the face below the midpoint of the area between the antennae
and a white patch centered on the ventral side of the face with black scales and hairs
elsewhere. The stripe may be thin to broad and may or may not fuse with the ventral


554
concept based on the ventral hindwing color, since 1 have not seen a transition between
specimens with a white background color and the background depicted in the Howe
(1975) painting. However, J. Emmel et al. 1998a report a population in the Inyo
Mountains oflnyo County, California, which "combines characters of subspecies
"neumoegeni, acastus, and vallismortis, with a general tendency to resemble
neumoegeni". From the black and white figures (of "subspecies" waucoba) illustrated by
J. Emmel et al. 1998a, this phenotype does not appear to me to differ from the normal
range of variation of the acastus phenotype, but these photographs do not illustrate the
coloration characters mentioned in the description. In any event, there is no mention in
the description of any morphological discontinuity between Inyo County specimens and
the rest of the C. acastus continuum of variation, so their is no basis for subspecific
recognition under my own subspecies concept. Finally, Austin (1998a) reports a
population from the Spring Mountains of Nevada (Clark and southern Nye Counties)
characterized relative to the nominate acastus phenotype by "larger in size than
nominotypical acastus with a more orange less yellowish aspect, broader black marks
and less basal black on the dorsal hindwing. On both surfaces...less contrast between the
paler and darker orange areas...pale areas of the ventral hindwing are deeper yellow...".
Austin (1998a) also reports the first three characters distinguish this phenotype from the
vallismortis phenotype, in addition to "on the ventral surface...a heavier black pattern
distally, a deeper orange color...and the basal pattern of the hindwing is set off by thicker
black lines". Judging from Austin's (1998a) figures 13-16, the extent of black markings
and the degree of contrast are well within the continuum of variation which exists for C.
acastus, and the wing span measurements reported based on forewing length (19.5-23mm


110
be justified. The shape of the valve in Gnathotriche is like that of Phycioditi in the sense
that the valvae taper posteriorly, and unlike D. didynia which has elongate posterior
pointed projections off of the ventral and dorsal posterior comers of non tapering valvae.
The inner valve process (harpe) of D. didyma is large, flattened, with several projections,
as opposed to an entire, smaller process in those Gnathotriche where the process is not
vestigial. Consequently, I found no justification for coding any characters of the inner
valve process or valve shape as the same state for D. didyma and Gnathotriche.
Since phylogenetic evidence indicates Gnathotriche forms the sister clade to
Phycioditi, retaining it in the Melitaeiti would require a Melitaeini classification scheme
with only two subtribes, the Euphydryiti and Melitaeiti. Alternatively, placing it in the
Phycioditi would change the concept of a valid higher taxon which has been in use for
several decades. In my opinion the best choice is to transfer Gnathotriche from the
Melitaeiti to its own subtribe, in order to produce a natural classification scheme that
preserves longstanding concepts of valid higher taxa. Also, the evidence for the
monophyly of Phycioditi and Gnathotrichiti individually is somewhat stronger than the
evidence for their combined monophyly, although evidence of their combined
monophyly is also well supported.
The phylogenetic analysis presented enables a reevaluation of Higgins (1981)
classification of placing G. exclamationis (Kollar), G. sodialis Staudinger, G. mundina
(Druce), and G. steinii (Dewitz) in two separate genera, with the former two taxa placed
in Gnathotriche and the latter two in Gnathotrusia. Higgins (1981) referred to G. steinii
as G. epione (Godman & Salvin), however Lamas (1989) presents evidence that epione is
a junior synonym o steinii. While I have unfortunately been unable to examine a


279
fusion or constriction of the black border, respectively. Whether these light areas are
continuous, separated, or if some of them are entirely filled in with black is an
intraspecifically variable feature, although if this variation could be quantified the means
would probably differ significantly between some taxa. The color of light scaling in the
middle of this patch is always identical to that of the middle of the discal cell. As with
the discal spot, this feature may be vary from sharp to obscure within taxa that vary from
light scaling to greatly suffused with black within the discal cell.
88. A patch of the symmetry system of wing pattern elements located within the discal
cell (formed from the fusion of the distal band of the basal symmetry system and the
basal band of the central symmetry system as noted above) distal to the basal part of the
cell and basal to the area where the discal cell is located, if present.
0=Present.
l=Absent.
2=Appearing present on the dorsal surface, but the dark symmetry system bands
are absent ventrally where a patch of darker (or indistinct) orange scales (relative
to the background) occurs in the equivalent position.
@*A=0&2: Varies along a continuum between state 0 and state 2.
I code ? for Chlosyne lacinia and Chlosyne californica. The presence or absence of the
primitive feature is difficult to definitively interpret in form saundersoni of C. lacinia
and in some specimens of C. californica. The other forms of C. lacinia and most
specimens of C. californica lack any trace of the Nymphalid ground plan black bands or a
patch formed by them in the forewing discal cell, as this part of the wing is all black.


796
Wright, W. G. 1906. The Butterflies of the West Coast of the United States.
Privately published.
790


725
specimens which were morphologically ambiguous as to representing subspecies browni
or montana. Females but not males have a more elongate forewing than females of the
other subspecies, while males have a more elongate forewing than subspecies mazarum
and riobalensis but not browni and rosita.
Further Description: The dorsal forewing frequently (but not always) has the median
band patches in cells M2-R5 distinctly enlarged relative to the other median band
patches, which is also the case in subspecies mazarum and riobalensis, but not in browni
or rosita. When these median band patches are not relatively large it is because the
posterior median band patches are large as well, while the sections from cells M2 to R5
are always prominent. Also, the dorsal forewing postmedian dots and white patches
basal to the median band are often (but not always) reduced and/or with some markings
absent on the dorsal wing surface such that the median band really stands out relative to
other forewing markings. This pattern also occurs to the same degree in most specimens
examined of subspecies riobalensis, but I have never seen it in subspecies browni. The
dorsal hindwing usually contains primarily black in the basal area basal to the position of
the distal symmetry system bands such that the basal light patch looks like a transverse
band, however there are some specimens with the light basal patch is the same size and
shape as in subspecies browni and rosita (the extremes of a continuum of variation). The
dorsal hindwing basal patch is light orange as in subspecies browni, and sometimes even
lighter, but never approaches the yellow/cream color present in subspecies mazarum.
The basal and sometimes the distal bands of the central symmetry system are (always?)
visible dorsally in the discal cell in those specimens where this position is not completely
black. All but one specimen examined have the dorsal hindwing median band extremely


563
Some Mexican records can not be located, because the label gives a the name of a city
which is not unique to one state: Santa Rosa (Chihuahua or Baja California Sur, I suspect
the former based on other distributional records), San Marcos (Coahuila or Guerrero),
Cordoba (Durango or Veracruz). Mexican localities include 80 miles NW of Mexico
City, Chihuahua (Santa Rosa), Hidalgo (Jacala, 4,000'; Jacala, 5200'), Tamaulipas
(Victoria), Nuevo Leon (20 miles N Monterrey), San Luis Potosi (Valles), Veracruz
(Region de Cordoba, Tierra Blanca), Mexico (Atizapan (near the north border of the
Federal District)) and the southeastern most record, Campeche (Lerma). Table 12
includes collection data for C. defmita from the FSCA and AM but in the case of the
NMNH primarily only records south of the U. S. border. Stanford and Opler (1993)
indicate additional records for southern New Mexico and the Mexican states of Sonora
and Coahuila. Higgins (1960) reports a record from Arizona with no further data, and
reports records from the Mexican states of Veracruz and Durango (Durango City). The
records cover a wide geographic area in Mexico, yet Mexican specimens are fairly poorly
represented in collections I have examined. Mexican records of specimens examined are
from the NMNH, FSCA, and AM, while Texas records are from the FSCA and AM
(many TX specimens are in the NMNH for which specific locality data was not
recorded). There is one specimen labeled Argentina in the Allyn Museum (Entre Rios
Salto Grande, 7.V.76, D. W. Jenkins) but since the next closest record is from Lerma,
Mexico, I must view this record with suspicion.
Species Delimitation: Phylogenetic evidence indicates C. defmita has no sister species.
There are gaps in pattern characters between C. defmita and all other Chlosyne. Higgins
(1960) expressed reluctance to include C. defmita and C. endeis as different species;


612
included ten subspecies (leanira, nebularum, obsoleta, oregonensis, wrightii, elegans,
austrima, basinensis, and flavodorsalis) and the other five names as synonyms. Austin
and Smith (1998b) explicitly state that integrades occur between seven of them. Scott
(1986) ranked fulvia, alma, and leanira as three subspecies of the specific taxon leanira.
Based on the long series of C. leanira leanira in the National Museum of Natural
History, the figures and descriptions in Austin and Smith (1998b) and Priestaf and J.
Emmel (1998), I find that my concept of C. leanira leanira includes a continuum of
variation with no gaps in the variation of any morphological character; hence I do not
divide this continuum into additional subspecies taxa. However, it is clear that the
continuum of variation is not a simple uniform geographic cline, but considerably more
complex. Austin and Smith (1998b) give a detailed and meticulous account of the
geographic variation within my concept of subspecies leanira (=their concept of species
leanira), including a distributional map with many specific localities. Unfortunately,
while the NMNH collection appears to me to exhibit the full range of variation for
subspecies leanira, many of the California specimens are severely lacking in specific
locality data, and many specimens have associated localities like "Calif.", "Middle Calif',
"Southern Calif.", "Los Angeles" (the county?, the city?, the closest major city?), etc. As
a result, I defer to Austin and Smith (1998b) for the most detailed account available of
the local geographic variation of C. leanira leanira. I have summarized the branching of
the continuum of variation based on information in Austin and Smith (1998b) combined
with my own observations in Figure 343, including the ten names Austin and Smith
(1998b) included as subspecies to refer to points along the continuum of variation. The
names in the diagram are used to indicate specific points on the continuum based on their


759
will be invaluable. Substantial changes are proposed for the classification of the
Melitaeini. In previous works, taxa had been delimitated highly arbitrarily, based on
personal opinions of degrees of similarity and differences. The evidence acquired from
this work strongly indicates such an approach resulted in the delimitation of a number of
artificial taxonomic concepts within the Melitaeini. The tools of modem systematics
employed by this work now enable one to treat the question of what constitutes a valid
taxon as a testable hypothesis.
Aside from the contributions to the systematics and evolution of the Meltaeini, I
believe a number of the results of several investigations presented in this work, based on
the Chlosyniti and Poladryiti as a case study, have significant implications for the field of
Lepidoptera systematics in general and/or systematic biology as a whole. With regard to
the issue of successively versus equally weighting characters, in this case study, based on
comparisons of congruence between independent data sets and with independent versus
combined data sets, equally weighting characters far outperformed successively
weighting characters. Evidence strongly suggests the added resolution obtained from
successive weighting of pattern characters alone was achieved at the expense of accuracy.
Comparisons of the weights assigned to particular character states indicates that all
characters are not of equal value to phylogeny construction nor were characters with
higher rescaled consistency indices proportionately better. The effectiveness of a model
proposed based on successively weighting character states rather than characters is a
worthy subject of future investigation.
Another interesting observation is that no evidence was obtained consistent with
the hypotheses that data sets with greater homoplasy, and thus lower tree statistics, are


151
Figures 40-42. Male genitalia capsule of Cinclidia phoebe. 40: Ventral. 41: Dorsal.
42: Lateral.


192


230
extensions in state 3 below except narrower and longer with only a small space of
lightly sclerotized tissue between), anterior to this area as in state 0 (Figure 97).
2=An elaborate spear-shaped design of lightly sclerotized tissue in the middle
bordered by darkly sclerotized tissue on the sides, with the light area anteriorly
gradually tapering to a sharp point. At the posterior end of the lightly sclerotized
triangular tapering portion the sides curve outward as posterior-lateral pointed
extensions such that only an extremely narrow amount of heavily sclerotized
tissue can be seen beyond their outer most point in ventral view. Posterior to this
area the outer border of the lightly sclerotized tissue occurs as an outward bulge
with a slight posterior slant. Between this bulge and the triangular tapering area
are two (one on each side) inward extensions of sclerotized tissue with squared
off apices (Figures 213-233).
3=Lightly sclerotized area ovoid, with the darkly sclerotized area reduced to an
extremely narrow band on each side at one point, and expanding anterior and
posterior to this point (Figures 127,129,131 & 133).
4=Posteriorly lightly sclerotized all the way across, anterior to this area the
darkly sclerotized area begins along the sides, widening until extending all the
way across forming a concave border (Figure 104).
With regard to the state 3 taxon Microtia elva, there is some intrataxon variability as to
the posterior expansion of the darkly sclerotized area. It is very slight in some specimens
(Figure 127), and more prominent in others.
10. The anterior extent of the lightly sclerotized area on the ventral posterior side of the
phallus:


374
I argue that if the assumption that all characters are of equal value to constructing
a phylogenetic hypothesis is true, and if the most parsimonious interpretation of the data
set constitutes the best evolutionary hypothesis, then there should be little or no character
conflict on the most parsimonious trees obtained. In other words, the number of extra
steps required by individual characters on the most parsimonious trees should be similar
(if the assumption of the model is a good approximation) or identical (if the assumption
of the model is exact). However, as indicated by Table 8, individual characters require
between 0 and 5 extra steps (the minimum number of steps is the number of states minus
one) on the strict consensus tree, and there is considerable variability in the number of
extra steps required by individual characters. If a most parsimonious tree is correct (or if
a consensus of multiple equally parsimonious trees is correct), any character that requires
more than the minimum number of steps on that tree is inherently unequal in its value for
reconstructing a phylogeny relative to a character which requires the minimum number of
steps, because the most parsimonious arrangement of that character can only be achieved
on an incorrect phylogenetic tree. Siebert (1992) states that there is no obvious
justification for an implication that some synapomorphies are of greater value than
others, while he also states that homoplasy occurs in almost all real data sets. Since
homoplasy occurs in almost all real data sets, how can all synapomorphies be of equal
value, when they cannot all be mapped on the same tree requiring the minimum number
of steps, regardless of what tree actually is correct? I argue that the very fact that
homoplasy occurs in almost all real data sets provides obvious justification that some
characters are more valuable than others for phylogeny construction. Since in the
Chlosyniti/Poladryiti data set of 139 parsimony informative characters (1) includes 72


690
band of C. hippodrome, absent in C. melanarge, is not to be confused with the
postmedian band of C. melanarge as it occurs basal to the white postmedian dots.
A very reliable character for separating C. hippodrome from the gloriosa
phenotype of C.janais is that C. hippodrome has a distinct red-orange patch in the basal
part of cell C on the ventral forewing, which is never present in any form of C. janais.
The gloriosa phenotype of C.janais has tiny postmedian dots, if any, in cells CuA2-M3,
while the dots are often but not always larger in C hippodrome. The red median band of
C. hippodrome is always divided into sections separated by black along the veins, while
in most specimens of the gloriosa phenotype of C.janais the sections of the band are
continuous and not separated by black. The ventral hindwing basal area of the gloriosa
phenotype of C.janais usually has prominent areas of cream color, whereas in C.
hippodrome this area is often solid black and has at most three or four small patches of
cream color. The male genitalia are very obviously different, with perhaps the most
obvious difference being the posterior valve process, which is short and rounded off in C.
hippodrome (Figure 202) and very long and pointed in C.janais (Figure 207).
The quehtala phenotype of C. lacinia is quite similar to C. hippodrome on the
dorsal surface but markedly different ventrally, although C. lacinia is highly variable
ventrally. Perhaps the most obvious consistent difference is C. lacinia always has an
orange patch in ventral hindwing cell CuA2 extending distal of the postmedian dot,
which is never the case in C. hippodrome.
The male genitalia of C. hippodrome are unique, and share the most
synapomorphies with three taxa which are markedly different in wing pattern: C.
ehrenbergi, C. gaudealis, and C. narva. The differences between C. hippodrome and C.


77
pouch is formed by an extra process, the abdominal stemite is attached to the base of the
lamella antevaginallis plate, and the process is free of the stemite throughout its length.
The out group taxon Catacroptera cloanthe has a pouch formed by a extension of
the abdominal tergite over the lamella antevaginallis. The posterior edge of the pouch
formed by the stemite is at the anterior edge of the ductus bursae opening. The lamella
antevaginallis itself is characteristic of state 0.
101. Given state 2 of the preceding character, the edge of the lamella postvaginallis with
respect to the lamella antevaginallis at the distal edge of the pouch:
l=Extended nearly even with the lamella antevaginallis (Fig. 264).
2=Extending well posterior of the lamella antevaginallis (Figs. 262-263 & 265-
266).
While independent evidence suggests state 2 of the preceding character is a terminal
derived state, taxa lacking state 2 of the preceding character are coded ? because this
character is not necessarily dependent on having state 2 for the preceding character.
However, taxa with the lamella pouch can be readily assigned to discrete states for this
character, which is not the case for taxa with the lamellae divergent.
102. An extension of the outer edges of the lamellae antevaginallis beyond the plane of
the rest of the lamella antevaginallis on the side opposite the ductus bursae opening,
forming concave depressions on each side:
0=This area is fairly smooth, lacking such extensions.
l=The above extensions are present (Fig. 283).
103. Folding of the distal edge (farthest from the genital opening) of the lamella
antevaginallis:


68
l=This feature present (Fig. 105).
78. The dorsal side of the everted vesica with a prominent, three dimensional sclerotized
structure attached to the dorsal posterior edge of the phallus. The structure is entire with
an anteriorly directed lobe proximally and elongate and tapering distally, extending well
past the teeth-bearing plates of the everted vesica:
0=Absent.
l=Present (Figs. 47-48 & 55).
Female genitalia
Characters of the Corpus Bursae:
79. The presence of a sclerotized plate on the ventral posterior side of the corpus bursae.
l=Present (Figs. 261-265,267-272,274,276-284 & 286-292).
#2=Absent.
Note that in Mellicta asteria the plate is very lightly sclerotized relative to other state 1
taxa, and may require higher magnification and more light to see. Since I only examined
one specimen of this taxon, this feature may not be consistent.
80. For those taxa with character 79 state 1, extensions of plate on ventral posterior side
of corpus bursae:
l^Plate extends anteriorly in two, narrow extensions, one on each side of the
plate (Figs. 261-265,278-284 & 286-292).
2=Plate lacks a pair of anterior extensions (Figs. 267,269-272,274 & 276).
The extensions are atypically short in Mellicta alatauica\ however, the posterior arch of
the plate is very characteristic of other Mellicta species examined.


648
the frontoclypeus anterior-lateral of each antennal base well developed. Frontoclypeus
with predominately black scales and hairs with sparse white hairs usually present on the
ventral part of the face. Patch of orange scales absent at the edge of the inner anterior-
lateral margin of the eye. Black and white but no orange scales present in the collar
between the head and pronotum. Posterior edges of posterior abdominal segments with
bands of white scales on dorsal and lateral sides, black scales anterior to these bands.
Ventrally abdomen with two longitudinal parallel black stripes against a yellow-cream
background, usually appearing continuous between each segment. Tibia and tarsi of
prothoracic legs with orange scales dorsally and white scales and hairs ventrally. Femur
of meso- and metathoracic legs with predominately orange scales dorsally and black at
the proximal end and predominately black scales ventrally except orange at distal end.
Tibia covered with orange scales only (erodyle) or with a mix of black and orange scales
with black most numerous (poecile). Tarsi with a variable mix of orange and black
scales. Meso- and metapleurons where the femur tucks in with black scales and hairs.
Thorax ventral to the wings with dense black scales and hairs, except white ones also
mixed in in females of subspecies erodyle. Antennal shaft with a checkered pattern of
black and white scales without orange scales, club with a white patch on the outer lateral
side.
Narrow band of orange scales along the base of the forewing costal margin. Wing
fringes checkered black and white with white sections between where the veins reach the
wing margin. Basal part of ventral forewing cell C with a distinct patch of orange.
Discal spot present and solid black, not detectable against the black background dorsally
or ventrally in most specimens of subspecies erodyle. Forewing discal cell with a patch


659
Species Delimitation: Chlosyne eumeda has wing pattern discontinuities with all other
Chlosyne. Phylogenetic evidence indicates C. eumeda is not most closely related to C.
marina or C. melitaeoides which have been considered the same species by some authors,
and I have seen specimens of both C. marina and C. eumeda labeled from Acahuizotla,
Mexico. I am uncertain if C. hylaeus is really a distinct species from C. eumeda, but
specimens of both phenotypes are labeled "Mexeala" (T. Escalante specimens) and I have
seen no evidence of integredation. However, given the rarity of C. hylaeus in collections
the possibility that integrades do occur can hardly be discounted. Furthermore, I am
aware of only two minor wing pattern characters which consistently separate C. hylaeus
and C. eumeda, the presence/absence of a cell M3 postmedian dot and the presence of a
well developed versus extremely reduced hindwing M3 postmedian band patch. Both of
these characters are intraspecifically variable in other Chlosyne species. In addition,
there is weak phylogenetic evidence that C. hylaeus and C. eumeda are sister taxa (if they
are distinct). I would not be surprised if in the future intermediates between C. hylaeus
and C. eumeda are discovered, and subsequently C. hylaeus would need to be
synonymized with C. eumeda. However, there is no justification for this action based on
the limited evidence that I have available to date.
Chlosyne hylaeus (Godman and Salvin)
Synchloe hylaeus Godman and Salvin, 1894. Ann. Mag. nat. Hist. (6)14:97. Type
Locality: Mexico-Durango. Holotype: British Museum [Examined via photo
courtesy of G. Lamas].


CHAPTER 1
INTRODUCTION
Overview
The Melitaeini (Lepidoptera: Nymphalidae: Nymphalinae) include a group of
over 265 butterfly species with representatives in Nearctic, Palearctic, and Neotropical
regions with the greatest diversification in the last (Higgins 1960, Higgins 1981). The
New World representatives of this group have been the subject of several major revisions,
including Higgins (1960), Higgins (1978), and Higgins (1981). The classifications
resulting from these revisions were based primarily on similarities and differences in
adult morphology. However, previous to the present work, the Melitaeini had never been
investigated with the tools of cladistic analyses to construct an evolutionary hypothesis of
the natural relationships within the group.
My initial interest in the Melitaeini was based on an interest in conducting a
phylogenetic study of the genus Chlosyne Butler and its close relatives. What attracted
me to this genus included a great complexity in morphological variation associated with
pattern characters (an impression that was immediately evident examining specimens in
museum drawers), and as a strikingly patterned group of butterflies long popular among
collectors, the genus was particularly well represented in collections. The combination of
complex morphological variation and extensive material available for study led me to
believe (correctly) that the Chlosyne were a group well suited for pursuing my interests in
phylogenetic studies of Lepidoptera. Early in the project, I decided that to achieve my
1


518
with cream scaling in the center of the dot (less common, most C. nycteis have cream
scaling). Distal median line and submarginal band sections deeply arrow head shaped
and pointing basally. Ventral hindwing and forewing between submarginal and marginal
bands with white and brown scales and no orange band.
Variation: Chlosyne gorgone exhibits a smaller range of variation than C. harrisii and C.
nycteis. There is variation in the extent of black on the dorsal forewing but it does not
reach the either of the two extremes occurring in C. harrisii and C. nycteis. There is also
variation in how light or dark the ventral background color is, which also does not appear
to change geographically. 1 have not noted any geographic variation despite an extensive
range, and consequently recognize no subspecies.
Gatrelle (1998) favored recognition of two subspecies, gorgone of the coastal
plain of Georgia/South Carolina and carlotta of north Georgia and elsewhere. Gatrelle
(1998) states "Its ["subspecies" gorgone's] two most distinguishing features are the lack
of a white pupil in the dark submarginal spot [=postmedian dot] in cell M3, and the
restricted white chevrons [=area between submarginal band and distal parafocal
elements] on the margin of the ventral hind wing." He also stated of "subspecies"
carlotta that it "is generally lighter." I have not examined specimens of C. gorgone from
coastal Georgia and South Carolina (other than the two Figured by Gatrelle (1998)),
however the absence or presence of cream scaling inside the M3 postmedian dot is
variable in C. gorgone throughout the rest of its range, and the size of the white crescents
between the submarginal band and distal parafocal elements is likewise variable
throughout its range. How light the ventral surface appears is variable throughout the
range of C. gorgone, and can also appear to vary based on the condition of specimens.


83
Phycioditi clade. If all but two Phycioditi are deleted (or the P. tharos three species clade
plus any other Phycioditi taxon), only two most parsimonious trees are obtained (before
successive weighting CI=0.892, RI=0.970, RC=0.865; after successive weighting
CI=0.940, RI=0.983, RC=0.924) with no change in any of the other clades. Of 115
parsimony informative characters, 88 characters require only the minimum number of
steps on the most parsimonious trees, and 27 characters are homoplastic. Status of
individual characters is presented in Table 2.
The strict consensus tree supports the monophyly of Higgins (1981) Euphydryiti
and Phycioditi, but indicates Melitaeiti is paraphyletic. Higgins (1981) concepts of
Antillea, Higginsius, Atlantea, Eurodryas, and Hypodryas come out as monophyletic
groups. However Higgins (1981) concepts of the following genera come out as
nonmonophyletic (*) or make a different genus paraphyletic(#): Texola*, Dymasia#,
Microtia#, Gnathotriche*, Gnathotrusia#, Occidryas*, and Euphydryas#. As noted,
insufficient taxa were examined to test Higgins (1981) concepts of Phycioditi genera or
Melitaea, Mellicta, Cinclidia, and Didymaeformia, however the taxa in the latter four
genera have grouped together with the limited sample examined. The Chlosyne group
including Chlosyne, Thessalia, Charidryas, and Anemaca, comes out as a monophyletic
group.
The shortest possible trees obtainable from successive weighting with the
representative out group method were obtained from the second successive weighting
analysis. Twelve equally parsimonious trees were obtained, with a consistency index of
0.926, retention index of 0.982, and rescaled consistency index of 0.910. The topology of
the strict consensus tree of successively weighted characters (Figure 294) is identical to


793
Hall, G. C. 1928-1930. A revision of the genus Phyciodes Hubn. (Lepidoptera:
Nymphalidae). Bull. Hill. Mus., Witley (as supplements to vols 2-4), pp. 1-44
(1928), 45-170 (1929), 171-205 (1930).
Harris, L. Jr. Butterflies of Georgia. University of Oklahoma Press, Oklahoma.
Harvey, D. J. 1991. Higher classification of the Nymphalidae. Appendix B in The
Development and Evolution of Butterfly Wing Patterns by H. Frederik Nijhout.
Smithsonian Institution.
Hawks, D., and J. F. Emmel. 1998. New subspecies of Cercyonis sthenele
(Boisduval) (Lepidoptera: Satyridae) and Chlosyne (Charidryas) gabbii (Behr)
(Nymphalidae) from the California islands. In Systematics of Western North
American Butterflies, Thomas C. Emmel, Editor. Mariposa Press. Gainesville,
FL. Pp. 321-326.
Hemming, F. 1967. The generic names of butterflies and their type-species
(Lepidoptera, Rhopalocera). Bulletin of the British Museum (Natural History),
Entomology, Supplement 9: 1-509.
Higgins, L. G. 1960. A revision of the Melitaeine genus Chlosyne and allied species
(Lepidoptera: Nymphalidae). Trans. R. Ent. Soc. Lond., 112: 381-467.
Higgins, L. G. 1978. A revision of the genus Euphydryas Scudder (Lepidoptera:
Nymphalidae). Entomologists Gazzette, 29: 109-115.
Higgins, L. G. 1981. A revision of Phyciodes and related genera, with a review of
the classification of the Melitaeinae. Bulletin of the British Museum (Natural
History), Entomology, 43: 77-243.
Howe, W. H. 1975. The butterflies of North America. Doubleday, New York.
International Commission on Zoological Nomenclature (ICZN). 1999. International
Code of Zoological Nomenclature. 4th Edition. London, International Trust for
Zoological Nomenclature.
Kitching, I. J. 1992a. The determination of character polarity. In Cladistics: A
Practical Course in Systematics by Peter J. Forey et al. Oxford University Press,
New York.
Kitching, I. J. 1992b. Tree-building techniques. In Cladistics: A Practical Course in
Systematics by Peter J. Forey et al. Oxford University Press, New York.
Lamas, G. 1989. The neotropical butterflies described by Dewitz in 1877 (Lepidoptera,
Rhopalocera). Zeitschrift fur Entomologie, Band 10, Heft 17, 273-279.
790


229
Taxa lacking character 4 state 1 are coded 0.
7. In ventral aspect, the sclerotization pattern of the posterior dorsal phallus extension
(states 0 versus 1 are best distinguished at 50X magnification, except for relatively large
specimens):
#0=Unifonnly sclerotized, except may appear darker on the sides where they are
folded over (Figures 83,88,92, & 97-98).
l=Sides and middle darkest (the sides may be as dark or darker than the middle)
with the sides and middle separated by an area of lightly sclerotized tissue of
variable width (Figure 228).
2=A triangle within a triangle pattern, with an lightly sclerotized arrow head
pattern in-between (Figure 104).
Since no taxa with different states for character 4 share the same state for the above
character, this coding implies no assumption that the two states of character 4 are or are
not forms of a homologous structure.
8. Posterior dorsal end of phallus hinged (moves dorsally when the vesica is everted:
0=This feature absent.
l^This feature present (Figure 105).
9. Sclerotization pattern of lightly and more darkly sclerotized areas on the ventral side
of the phallus:
#0=A narrow band of dark sclerotization on each side, and lightly sclerotized in
the middle (Figures 88,92 & 98).
*#l=Posteriorly with the appearance of narrow squared off inward extensions of
sclerotized tissue anterior to the teeth bearing plate (somewhat similar to the


381
example. Based on this case study, I find the errors associated with ambiguous
distributions of character states on a tree are minor compared to the error associated with
weighting character states based on the rescaled consistency indices of their character.
Instances of ambiguous state distributions are relatively far fewer than the abundant
examples where overall character quality is a poor reflection of character state quality
depicted in Table 11.
Another, and potentially far more serious source of weighting error (for either
characters themselves or their states) is exemplified by Figures 334-335, which represent
strict consensus trees (for analyses exclusively based on pattern characters) derived from
equally and successively weighting characters. In the strict consensus tree of equally
weighted characters, the relationships between Polcidryas, Atlantea, Higginsius, Antillea,
the Microtia taxa (other than the (comeara,anmalas) clade), and Chlosyne are
unresolved. While genitalic characters provide robust support of these relationships
(Figures 324 & 325 with numbers 1-52 representing genitalic characters), pattern
characters lack sufficient information to resolve these relationships (there are no
universal pattern character state synapomorphies supporting these relationships, although
there are some assemblages of unique derived nonhomoplastic states of multistate
characters). When all characters are included in the analysis, the support for these
relationships is robust, and the most parsimonious trees identify particular pattern
characters as being homoplastic. However, due to the poor resolution on this part of the
tree when only pattern characters are included in the analysis, these patterns of
homoplasy are not recognized when characters are reweighted. After pattern characters
are successively weighted in the analysis with only pattern characters, the result is a


775
I E$k


644
melitaeoides are two species or two subspecies is arbitrary, but given the current absence
of evidence that the two are monophlyetic, there is no justification for including them as
members of one taxon.
Chlosyne melitaeoides (C. and R. Felder)
Figures 142,165,186,221 & 242
Synchloe melitaeoides C. and R. Felder, 1867. Reise Novara. 3:396. Type Locality:
Mexico. Holotype: British Museum [Examined via photo courtesy of G. Lamas],
Diagnosis: Chlosyne melitaeoides may be confused with C. marina, C. eumeda, C.
erodyle poecile, and C. endeis pardelina. It may be separated from C. marina by the
characters given in the diagnosis for that taxon. The same characters noted under C.
endeis endeis which separate that taxon from C. marina also separate any specimen of C.
endeis from C. melitaeoides.
C. melitaeoides differs from C. eumeda by having distinctly smaller forewing
yellow median band patches, and in particular the patches of C. eumeda in cells CuA2
and CuAl are broad and aligned basally and distally which is not the case in C.
melitaeoides (exceptions for C. eumeda have been found in two melanic specimens
which have reduced forewing and hindwing bands but otherwise have typical of C.
eumeda characters). The dorsal forewing postmedian dots are sharp white in C.
melitaeoides versus cream to pale yellow in C. eumeda. There is only a thin amount of
black basal to the ventral hindwing postmedian band patches in C. melitaeoides, while


162
Figs. 86-88. Male genitalia of Higginsiusfasciatus. 86: Capsule (dorsal). 87: Capsule (ventral).88: Phallus
(ventral). Figs. 89-92. Male genitalia of Higginsius miriam. 89: Capsule (ventral). 90: Capsule (dorsal).
91: Capsule (lateral). 92: Phallus (ventral). Fig 93: Male genitalia capsule of Poladryas minuta (posterior inverted)


1
98
in a variable pattern but never cover the entire corpus bursae. Ductus bursae short or
vestigial. Ostium bursae present or absent. Lamella postvaginallis and lamella
antevaginallis present, although the latter sometimes greatly reduced, and occurring in
various forms.
Euphydryiti Higgins
Synapomorphies from binary characters: Inner ventral wall of valvae twisted roughly 90
degrees and lobed, forming a pocket shaped roughly like a half circle containing spine
like setae on its posterior surface (Fig. 3). Saccus reduced to a very narrow bridge along
the ventral anterior side of the genitalia capsule (Fig. 2). Minute granulse patches
densely covering projections on the tegumen (Figs. 19-24). Proximal area of everted
vesica densely encircled with minute granulse patches conspicuous at 50X
magnification with good lighting (Figs. 27-28 & 33).
Terminal derived states of multistate characters: Inner valve process forked (Figs. 6-14).
Ventral valve opening formed between where the valve plate folds around completely
visible in lateral view (Fig. 15). Paired projections on an otherwise simple tegumen
connected by a thin bridge of tegumen between them (Figs. 16-23). Plate on ventral
corpus bursae disjunct, separated by membranous tissue along the ventral midline. Teeth
on the inner surface of corpus bursae encircle the corpus bursae in a band of similar width
(Fig. 261).
Further description: Males: Valvae lacking projections or processes other than the inner
valvae projections characteristic of all Melitaeini. Sockets for setae on inner sides of
valvae visible in ventral view. A patch of short setae is present on the dorsal half of the
posterior lateral sides of the valvae anterior to the area with dense hair like posterior


55
51. For those taxa with character 49 state 1, the presence of an additional pointed
projection below the apex on the inner sides of the tegumen projections:
l=Absent.
2=Present (Figs. 19 & 21).
52. For those taxa with character 49 state 2, further characterization of the projections off
of the tegumen:
l=About as long as length of tegumen anterior to the projections or slightly
shorter, for most of length at most slightly tapering; however, apically projections
curve inward a short distance and taper abruptly to a point (Fig. 35).
2=About as long as length of tegumen anterior to the projections or slightly
shorter, outer side convex and inner side straighter and slightly concave, apex
pointed (Fig. 2).
3=Small, triangular, and much shorter than the length of the tegumen anterior to
the projections, apex pointed (Figs. 86 & 90).
4=Weakly triangular and much shorter than the length of the tegumen anterior to
the projections, outer side somewhat convex, inner side slightly convex
proximally and straight distally, apex blunt (Fig. 80).
Independent evidence from other characters suggests state 2 of the third preceding
character independently arose twice; however, there is no overlap of the above states
between the two clades where this state arose. Consequently, taxa lacking state 2 of the
third preceding character can be coded 0 without risk of doubly weighting an independent
acquisition.


Table 12 Continued
733
NMNH=Nat. Mus, of Natural Hist.-Smithsonian Inst.: FSCA=FL St. Collection of Arthropods; AM=Allyn Mus ; PC=Personal Collection; @=Judged by author as of dubious authenticity
Country
County (USA)
Locality (non US)
Locality (USA)
Locajity^ContinuedOTO
i USA)
USA
.178. Aster umbullatus
1.78. Aster umbullatus
Chlosyne herrisii
USA
Wisconsin
Jackson
Chlosyne harisli
USA
Wisconsin
Jackson
Chlosyne harisli
USA
Wisconsin
Jackson
Chlosyne harlsii
USA
Wisconsin
Jackson
Chlosyne harisii
USA
Wisconsin
Maathon
Chlosyne harisii
USA
Wisconsin
Outagamie
Chlosyne harisii
USA
Wisconsin
Chlosyne harisii
USA
Wisconsin
Outagamie
Chlosyne harisii
USA
Wisconsin
Outagamie
Chlosyne harisii
USA
Wisconsin
Outagamie
Chlosyne harisii
USA
Wisconsin
Outagamie
Chlosyne harisii
USA
Wisconsin
Outagamie
Chlosyne harisii
USA
Wisconsin
Outagamie
Chlosyne hahsii
USA
Wisconsin
Outagamie
Chlosyne nycteis (FSCA-part)
Chlosyne nycteis
USA
Georgia
Fulton
Chlosyne nycteis
USA
Kansas
Chlosyne nycteis
USA
Maryland
Harford
Chlosyne nycteis
USA
Minnesota
Cook
Chlosyne nycteis
USA
Missouri
Holt
Chlosyne nycteis
USA
Pennsylvania
Chesta
Chlosyne nycteis
USA
Texas
Hamilton
Chlosyne nycteis
USA
Texas
Hamilton
Chlosyne nycteis
USA
Wisconsin
Jackson
Chlosyne nycteis
USA
Wisconsin
Mamette
Chlosyne nycteis
USA
Wisconsin
Chlosyne nycteis
USA
Wisconsin
Chlosyne nycteis
USA
Wisconsin
Mainelte
Chlosyne nycteis
USA
Wisconsin
Mamette
Chlosyne nycteis
USA
Wisconsin
Mamette
Chlosyne nycteis
USA
Wisconsin
Chlosyne nycteis
USA
Wisconsin
Chlosyne nycteis
USA
Wisconsin
Mamette
Chlosyne nycteis
USA
Wisconsin
Outagamie
Chlosyne nycteis
USA
Wisconsin
Outagamie
Chlosyne nycteis
USA
Wisconsin
Outagamie
Chlosyne nycteis
USA
Wisconsin
Outagamie
Chlosyne nycteis
USA
Wisconsin
Outagamie
Chlosyne nycteis
USA
Wisconsin
Chlosyne nycteis
USA
Wisconsin
Outagamie
Outagamie
Chlosyne nycteis
USA
Wisconsin
Chlosyne nycteis
USA
Wisconsin
Winnebago
Scott Bog, ex larva VIII.22.77. Aster umbullatus
Whitefield
Whitefield
Brookline
[Bedford
Black Moshannon St Park
Black Moshannon St Park
Black Moshannon St Park
Black Moshannon St Part, ex I
Black Moshannon St Part, ex I
Black Moshannon St Part, ex I
Black Moshannon St. Park, ex I
Black Moshannon St. Part, ex I
Black Moshannon St Park
Black Moshannon St Park
I Black Moshannon Si Part, e> egg
Black Moshannon St. Part, ex egg I
Black Moshannon St Part, ex I
Tamarack
Baldwin Run Road. Delmar Twp 2100'
Head. Carpenter Hollow. Delmar Tnp 2100'
Head. Carpenter Hollow. Delmar Tnp 2100'
Head, Carpenter Hollow. Delmar Tnp 2100'
Head. Carpenter Hollow. Delmar Tnp 2100'
; Head. Elk Run. Covington Tnp.. 2600'
¡Sleepy Fool
Slippery Rock
: Spruce Knob Lake
Spruce Knob Lake
Armstrong Creek Bog, T 37N R 16E Sec 13/14. in bog
Armstrong Creek Bog. T 37N R 16E Sec 13/14, in bog
E Castle Mound Rd sedge meadow. T 21 N R 3W Sec 26
E C asile Mound Rd sedge meadow, T 21 N R 3W Sec 27
E Castle Mound Rd.. sedge meadow, T 21 N R 3W Sec 36
4 July 1974
13 June 1976
W J Kid
W JKiel
Jy 14 38
lvi.20.75
Em. VI. 11.78
VI 20.1978
Em. 21 Juna 1963~
982, Aster umBullalus Em 23 June 1983
[VI.21.75
VI 12 76
4 JuneJI977
Vl.4.1977 ~
24 June 1978
21 June 1966
20 June 1966
.301 J une 1966
30 July 1955
24 June 1966
VI-13-27
7 July 1979
VII-10-83
24 June 1995
25 June 1995
4 July 1993
4 July 1993
Fellen Timbers EnvironmenlBl Center. T 23N R
E Castle Mound Rd.. sedge meadow. T 21 N R 2W Sec 31
E Casile Mound Rd.. sedge meadow. T 21 N R 3W Sec 26
E Castle Mound Rd., sedge meadow, T 21 N R 3W Sec 27
Fellen Timbers Envtronmeniel Center. T 23N R 17E Sec 9. sedge m-
Fellen Timbers Environmental Center. T 23N R 17E Sec 9. sedge rrv
Timbers Environments! Center, T 23N R 17E Sec 9. sedge rrv
7E Sec 9, edge tnt
Fellen Timbers Envtronmentel Center,
Fellen Timbers Envtronmeniel Ct
'E Sec 9. sedge meadow
Timbers Envttonmenlal Center. T 23N R 17E Sec 9. se
Fellen Timbers Environmental Center. T 23N R 17E Sec 9, se
Fellen Timbers Envtronmeniel Center, T 23N R 17E Sec 9, se
4 July 1993
5 July 1993
5 July 1993
] 21 June 76
9 June 1991
10 June 1991
12 June 1992
.20 June 1992
j 16 June 1993
17 June 1993
112 June 1995
114 June 1995
115 June 1995
G F Patterson
| FSCA
¡FSCA
¡FSCA
>scs ;
FSCA
iFSCA
FSCA
FSCA
FSCA
FSCA
FSCA
.FSCA
FSCA
FSCA
FSCA
[FSCA
FSCA
FSCA
FSCA
[FSCA
FSCA
FSCA
FSCA
FSCA
FSCA
FSCA
1 FSCA
FSCA
1PSCA
FSCA
FSCA
NMNH
FSCA
.NMNH
;PC
_ Hugo L. Kons Jr.
Hugo L. Kons Jr.
Hugo L Kons Jr.
Hugo L Kons Jr,
Hugo L Kons Jr.
I Hugo L Kons Jt.
R A Rahn
Challa River
Eureka
- |
; 7
Near McCormick Wildlife Area, Loe
Nottingham
E Castle Mound Rd., moist shri
' Shnne Rd. xenc-mesic loresl edg op
a. T 2INR3W S.
boiTens. T 37N R
a. 1968 I
a 1968 I
I Hugo L. Kon Jr.
Hugo L Kons Jr.
Hugo lT Kons Jr.
I Hugo L Kons Jr.
Hugo L Kons Jr
| Hugo L. Kons Jr.
! Hugo L- Kons Jr
HugoL. Kons Jf.
Hugo L Kons Jr
Abner A Towers
lELTodd
IR S Simmons
|J H Masters
J R Beitzman
FDFee
Roy O Kendall
Roy O Kendall
Hugo L- Kons Jr.
HugoL Kon Jr.
Chiosyne gorgone
Chlosyne got gone
Chlosyne gorgone
h-t
Chlosyne gorgone
Chlosyne gorgone
Chlosyne gorgone
Chtosyn. gorgone
USA
USA
USA
USA
USA
USA
USA
USA
Chlosyne gorgone
USA
Wisconsin
Outagamie
Chlosyne gorgone
USA
Wisconsin
Outagamie
Chlosyne gorgone
USA
Wisconsin
USA
Wisconsin
Chlosyne gorgone
USA
Wisconsin
Outagamie
Chlosyne gorgone
USA
Wisconsin
Outagamie
Chlosyne gorgone
USA
Wisconsin
Outagamie
Chlosyne gorgone
USA
Wisconsin
: Outagamie
Chlosyne gorgone
USA
Wisconsin
Chlosyne gorgone
USA
Wisconsin
Chlosyne qorqone
USA
Wisconsin
Portaqe
ea barrens. T 37N R 17E Sec 14
a, T 37N R 17E Sec 11
a, T37N R 17E Sac 13
Shnnt Rd, xanc-rwalc loraal adyopan artaatbatrana, T 37N R 17E Sec 24
| Shnne Rd. xenc-maaic lores! edg open areas barrena, T 37N R 17E Sac 13
| Shnne Rd, xenc-meetc loraal edg/open areaebarrera. T 37N R 17E Sec 14
Dunbar Barrens, oak-pine barrens/forest edg T37N R I8E Sec 21
ie Co. Rec Trail between Medina and Wjnnebaqo Co.
ill Nature Center, xenc hill side/top. T 22N R 15 E
¡II Nature Center, xenc hill side/top, T 22N RISE
ill Natura Center, xatic hill ilde/lop. T 22N R 15 E
ill Nature Center, xaric hkl sldeftop. T 22N RISE
ill Nature Center, xatic hiH aide/top, T 22N R 15 E
III Natura Center, xerlc hill aide/top. T 22N R 15 E
¡II Nature Center, xeric hill side/top. T 22N R 15 E
Winnebago Co Rec Trail, 0-0 5 miles S of Outagamie Co
Red Rocks Park
I Pine ME
Medicine Lodge, ex larva
3 miles N1
[Vic Irving
S, Augustine Baykin Sprgs.
a mosaic. T 42N R 14W Sec 1
Murall Bluff Prarie. dry prairie on bluff, T 3N R 8E Sec 25
Mosquito Hill Nature Center, prarie planting. T 22N R 15E
Mosquito Hill Nalure Center, prarie planting, T 22N R 15E
Mosquito Hill Nalure Center, prairie planting, T 22N R 15E
Mosquito Hill Nature Center, prairie planting, T 22N R 15E
Mosquito Hill Nalure Center, prarie planting. T 22N R 15E
Mosquito Hill Nalure Center, prarie planling, T 22N R 15E
Mosquito Hill Nature Center, prarie planting. T 22N R 15E
Fallen Timbers Env Cntr. xeric meadow, T 23N R I 7E Sec 9
Fallen Timbers Env Cntr.. xenc meadow. T 23N R I7E Sec 9
Yadadj lo Plaman Co Pak. ex larva on Gloriosa Dasy
Emmons Crk Public Hum, Ar., oak-pine barens opening


447
8. Wing fringes cream colored (sometimes with brown scales present also), ventral
hindwing with broad black scaling along the veins and cream scaling in between with a
postmedian band and distal and basal parafocal elements completely absent
Chlosyne ehrenbergi
8. Wing fringes checkered black and white, if ventral hindwing not completely black
(other than for postmedian dots) distally, some form of a postmedian band is present ...9
9. Vertex lacking a white patch (if white scales are present they are scattered and not
organized into a patch), submarginal band distinct against the background, either fairly
straight or forming arrow heads, hindwing postmedian dots present in cells CuA2-Ml
within a postmedian band of colored patches and composed of black and/or cream
colored scales (usually both) 10
9. Vertex with a distinct white patch reaching its posterior edge, submarginal band not
distinct against the background (position of submarginal band to distal edge of wing is
solid black), hindwing postmedian dots differ from above 20
10. Hindwing postmedian dots composed predominately of black scales with a complete
solid black border around any dots with cream scales in the center (except for the dot in
cell M3 in some C. harrisii), dorsal hindwing between the distal parafocal element and
submarginal band predominately black with a row of diffuse small light colored crescents
with the basal sides convex and the distal sides concave, lateral sides of thorax with pale
orange hairs absent 11
10. Hindwing postmedian dots composed predominately of cream scales and lacking a
complete solid black border, dorsal hindwing between the distal parafocal element and
submarginal band with a well developed row of light patches with the basal sides convex


95
the ancestral node to Phycioditi will be ambiguous. However, there is considerable
variation in wing pattern among many Phycioditi where homologies in wing pattern
elements are evident.
Revised Higher Classification of the Melitaeini
*Euphydryiti Higgins (Holarctic)
*Euphydryas Scudder (including Occidryas Higgins)
*Hypodryas Higgins
*Eurodryas Higgins
#Melitaeiti Tutt (Palearctic)
Mellicta Billberg
Melitaea Fabricius
Didymaeformia Verity
Cinclidia Hubner
#Phycioditi Higgins (Nearctic & Neotropical)
Phyciodes Hubner
Phystis Higgins
Anthanassa Scudder
Dagon Higgins
Telenassa Higgins
Ortilia Higgins
Tisona Higgins
Tegosa Higgins
Eresia Boisduval
Castilia Higgins
Janatella Higgins
Mazia Higgins
*Gnathotrichiti subtribe n. (Neotropical)
*Gnathotriche Felder & Felder (including Gnathotrusia Higgins)
**Poladryiti subtribe n. (Nearctic & Neotropical)
** Pol a dry as Bauer
**Atlantea Higgins
**Higginsius Hemming
**Chlosyniti subtribe n. (Nearctic & Neotropical)
**Antillea Higgins
**Microtia Bates (including Dymasia Higgins and Texola Higgins)
**Chlosyne Butler (including Thesscilici Scudder, Charidrycis Scudder, & Anemaca)
*=A phylogenetic hypothesis has been reconstructed for relationships within this higher
taxon.
**The evidence supporting this arrangement is in part provided by the phylogenetic
analysis in the following chapter or supplemented with analyses in the following chapter.


59
58. For those taxa with state 2 for the preceding character, the point of the posterior
extent of the more heavily sclerotized area:
l=Much closer to the posterior opening than the supersensory membrane (Figs.
49-51 & 53).
2=Closer to the supersensory membrane than the posterior opening (Figs. 26,30 &
104).
Taxa lacking state 2 for the preceding character are coded
59. The anterior extent of the lightly sclerotized area on the ventral phallus surface:
0=Reaching (and may surpass) the supersensory membrane (Figs. Figs. 26,30,49-
51,53,65,68,73,79,88,92,97-98,104,127,213-214,216-224).
1 terminating distinctly posterior to the supersensory membrane, but closer to
the supersensory membrane than the posterior phallus opening (Figs. 129,131,
133, 225-227,230,232-233).
2=Terminating distinctly posterior to the supersensory membrane, but closer to
the posterior phallus opening than the supersensory membrane (Fig. 53).
This character varies within the genus Chlosyne. 1 code Chlosyne as ? in this analysis
but code Chlosyne taxa individually for this character in the Chlosyniti data matrix.
60. Sclerotization pattern of the dorsal surface of the phallus:
l=The longitudinal middle is more heavily sclerotized relative to the sides, and
this sclerotized area tapers anterior to posterior (within this area the midline may
appear lighter than the sides) (Figs.29 & 31).
2=Uniformly sclerotized (the sides may appear darker in dorsal view due to their
proximity to sclerotized area of the phallus on the ventral side) (Figs. 69 & 77).


561
with cream colored postmedian dots inside. Hindwing postmedian dots limited to one
cream colored dot in cell M3, which is often diffuse and best seen with magnification in
northern specimens but may be large and conspicuous in southern ones where the M3
orange postmedian band patch is greatly reduced. Distal and basal median lines present
on the ventral hindwing surface, in the form of mostly continuous uneven lines. Dorsal
and ventral hindwing and forewing margin between the position of the marginal and
submarginal bands solid black. Dorsal hindwing and forewing patches comprising the
median band light orange to cream and concolorous. Dorsal hindwing between the
position of the submarginal band and distal parafocal element with sharp (the edges may
or may not be diffused with black) orange to cream patches of variable shape and size
against a black background. Prominent patches present in the same position ventrally
where they are light cream colored with the distal borders fairly straight or slightly
convex and the basal borders often strongly convex. Patches formed between the distal
and basal elements of both the basal and central symmetry systems on the ventral
hindwing coalesced together with the symmetry system black bands giving the
appearance of a continuous band. The bands formed between the bands of the distal and
basal symmetry system are at most slightly fused together and usually distinctly separate.
Symmetry systems bands are complete in all cells and black with orange between them,
although the orange between bands of the basal symmetry system tends to be sparse
beyond the costal cell in some southern specimens. Ventral hindwing background color
between the proximal band of the central symmetry system and distal band of the basal
symmetry system cream and concolorous with the rest of the background color (at most a


Table 1-Continued
136
Characters
Taxa
61
62 63 64 65
66 67 68 69 70
71
72 73 74 75
76 77
78 79 80
*Eurodryas aurinia
1
1
0
3 2
1 1
1 1
1
1
0 0 0
1
0 0
0 1
1
Eurodryas desfontaini
1
1
ot
3 2
1 1
1 1
1
1
0 0 0
1
0 0
0 1
1
*Hypodryas maturna
1
1
trf
3 2
1 2
0 1
1
1
0 0 0
1
0 0
0 1
1
Hypodryas intermedia
1
1
0
3 2
1 2
0 1
1
1
0 0 0
1
0 0
0 1
1
Hypodryas gillettii
1
1
0
3 2
1 2
0 1
1
1
0 0 0
1
oto
0 1
1
Hypodryas cynthia
1
1
0
3 2
1 2
0 1
1
1
0 0 0
1
0 0
0 1
1
Hypodryas iduna
1
1
of
3 2
1 2
0 1
1
1
0 0 0
1
0 0
0 1
1
*Occidryas anicia
1
1
0
3 2
1 2
0 1
1
1
o oto
1
0 0
0 1
1
Occidryas chalcedona
1
1
IT]
3 2
1 2
0 1
1
1
0 0 0
1
0 0
0 1
1
Occidryas colon
1
1
0
3 2
1 2
0 1
1
1
0 0 0
1
0 0
0 1
1
*Euphydryas phaeton
1
1
tot
3 2
1 2
0 1
1
1
0 0 0
1
0 0
0 1
1
Occidryas editha
1
1
0
3 2
1 2
0 1
1
1
0 0 0
1
0 0
0 1
1
Out Group Taxa
Colobura dirce
1
1
0
? ?
0 0
0 1
1
1
0 0 0
1
0 0
0
Kallima paralekta
1
c
0
? ?
0 0
0 1
1
1
0 0 0
1
0 0
0
Anartia jatrophe
1
1
0
? ?
0 0
0 1
1
1
0 0 0
1
0 0
0 2
0
Anartia amathea
1
1
0
? ?
0 0
0 1
1
1
0 0 0
1
0 0
0 2
0
Anartia fatima
1
1
0
? ?
o Vo
0 1
1
1
0 0 0
1
0 0
0 2
0
Anartia chrysopelea
1
1
0
? ?
oT o
0 1
1
1
0 0 0
1
0 0
0 2
0
Nymphalis polychloros
2
1
0
? ?
0 0
0 1
1
1
0 0 0
1
0 0
oT
Hypanartia lethe
1
c
0
? ?
0 0
0 1
1
1
0 0 0
1
0 0
0
Vanessa atlanta
1
c
0
? ?
0 0
0 1
1
1
0 0 0
1
0 0
0
Cynthia cardui
1
c
ot
? ?
0 0
0 1
1
1
0 0 0
1
0 0
0
Hypolimnas pandarus
1
c
OT
? 2
0 0
0 1
1
1
0 o |o
1
0 0
0
Amnosia decora
1
c
ot
? ?
0 0
0 1
1
1
0 0 0
1
0 ?
0
Araschnia levana
1
c
0
? 2
0 0
0 1
1
1
0 0 0
1
0 0
0
Araschnia prorsa
1
c
0
? 2
0 0
0 1
1
1
0 0 0
1
0 0
0
Agais urticae
1
c
0
? ?
0 0
0 1
1
1
0 0 0
1
0 0
0
Precis octavia
1
1
"ot
? ?
0t 0
0 1
1
1
0 0 0
1
0 0
TT
Catacroptera cloanthe
1
c
0
? 2
0 0
0 1
1
1
0 0 0
1
0 0
0 2
0
Inachis io
1
c
0
? ?
0 0
0 1
1
1
0 0 0
1
0 0
Junonia coenia
1
1
0
? 2
0 0
0 1
1
1
0 0 0
1
0 0
0 2
0
Polygonia c-aureum
1
c
0
? ?
0 0
0 1
1
1
0 0 0
1
0 0
ITT
Metamorpha elissa
1
c
0
? ?
0 0
0 1
1
1
0 0 0
1
0 0
Irr
Siproeta epaphus
1
1
0
? 2
0 0
0 1
1
1
0 0 0
1
0 0
ITT
Yoma sabina
1
c
0
? ?
0 0
0 1
1
1
0 0 0
1
0 0
oT
Doleschallia bisaltide
1
1
0
? 2
0 0
0 3
1
1
0 0 0
1
0 0
"oT
Salamis augustina
1
c
0
? 2
0 0
0 1
1
1
0 0 0
1
0 0
-ot
Napeocles jucunda
1
c
0
? 2
0 0
0 1
1
1
0 0 0
1
0 0
~ot
Rhinopalpa polynice
1
0
? 2
0 0
0 1
1
1
0 0 0
1
0 0
0
Vanessula milca
1
1
0
? 2
0 0
0 1
1
1
0 0 0
1
0 0
0
Antanartia delius
t
c
0
? ?
0 0
0 1
1
1
0 0 0
1
0 0
oT
Kaniska canace
1
c
0
? ?
0 0
0 1
1
1
0 0 0
1
0 0
0
Cumulative Out Group
1
1,c
0
? ?
0 0
0 1
1
1
0 0 0
1
0 0
0


405
Ratio of Posterior Valve Projection Length to Posterior Valve Length for Selected Taxa
0 2 4 6 8 10 12 14 16
1 pala
I gabbii.acasius.
whitncyi
A 3 hoffmani
X 4 harrissi
X 5 nr. harrissi
6nycteis
+ 7 gorgone
- 8 defhia
- 9 endeis
010 erodyle
11 eumeda
A12 marina
13 meliaeoides
14 melanarge
o 15 lacinia
Taxa
Figure 310: Ratio of posterior valve projection length to posterior valve length for selected Chlosyne.


548
report that some specimens from the mainland coast in Santa Barbara and San Luis
Obispo Counties (not seen by the author) are darker than nominate gabbii and "may
represent somewhat of an intermediate phenotype", and since the differences between the
typical gabbii phenotype and the illustrated and specimens assigned to the atrifasciata
phenotype are quite small, I interpret this as evidence that atrifasciata and gabbii
phenotypes represent ends of a continuum of variation in the size of the basal parafocal
elements, and consequently do not formally recognize the atrifasciata phenotype as a
subspecies.
Wing Span: S' 34-41mm. $:39-44mm.
Range: I have seen specimens only from southwest California, including Los Angeles
County (Los Angeles, Fullerton, San Fernando), San Diego County (Pulomas Mountains,
Spring Valley-La Pressa Canyon, San Diego), Riverside County (South of Hemet), Santa
Barbara County (Montecito), and Monterey County (Chews Ridge-most northemly
locality). Collection data for C. gabbii specimens from the NMNH and FSCA are
presented in Table 12. Scott (1986), Brown et al. (1992) and Stanford and Opler (1993)
also report C. gabbii from northern Baja California in Mexico. Stanford and Opler's
(1993) California county distributional map shows records of C. gabbii as far north as
Santa Clara and San Benito counties along the west central California coast and in all
counties south of these except for the extreme southeast county (Imperial County).
Species Delimitation: Based primarily on the hybridization experiments of J. Emmel
reported by Scott (1986) noted under C. whitneyi (above). Also, C. gabbii appears to
have a consistent wing pattern discontinuity with C. acastus, and has wing pattern
discontinuities with all other taxa in the C. palla clade.


695
Papilio iiarva Fabricius, 1793. Ent. Syst. Emend. 3(1):249. Type Locality: "Africa".
=Cethosia bonpland Latreille, 1809. in Humbolt, Voy. Reg. Equinox. Nouv. Cont.: 282.
Type Locality: Peru-Cuenca.
=Cethosia bonplandi Godart, 1819. Encyc. Meth. 9:245. Type Locality: Peru-Cuenca.
Diagnosis: Chlosyne narva is a distinctive taxon. The only Chlosyne with a similar wing
shape with elongate forewings so produced at the apex is C. gaudecdis, which differs
from C. narva in having the forewing discal cell with a red to red-orange patch or all
black. I have seen a specimen of C. narva confused with C. rosita, but in all subspecies
of C. rosita the dorsal forewing markings are only black and white, which is never the
case in C. narva.
The male genitalia of C. narva are most similar to C. gaudealis, but there are
consistent differences between them. The inner valve process is much straighter and
angled more posteriorly (forming approximately a 45 angle relative to a horizontal line
bisecting the genitalia) in ventral view in C. narva (Figure 152) relative to C. gaudealis
(Figure 153), where the concave curvature of the inner valve process is more evident and
the angle of the process with the horizontal is distinctly less than 45. Both taxa have the
inner valve process squared off apically, however the apex is somewhat more flared out
in C. gaudealis relative to C. narva in ventral view. In dorsal view, the inward curve of
the ventral valve process is somewhat less in C. narva (Figure 172) relative to C.
gaudealis (Figure 173). Also, in lateral view the anterior dorsal comer of the vinculum is
narrowly rounded off and forming a slightly obtuse angle in C. narva (Figure 174), while


Table 12 Continued
744
NMNH=Nal Mus of Natural Hist.-Smithsoman Inst FSCA=FL St Collection of Arthropods. AM=Allyn Mus PC=Personal Collection. ffl=Judged by author as of dubious authenticity
1
County (USA)
Locality (USA)
Taton
| Country
State/Pro v.
Locality (non US)
Locality Continued (non USA)
Date
Collector
Collectio
No
Chlosyne lacinia
USA
Arizona
7
¡EL Todd
NMNH
1
Chios yna lacinia
USA
Arizona
7
NMNH
7
Note NMNH California records with no form specified are form ciocate or croc ate/ad|uUix intermediate forms.
Chlosyne lacinia
USA California
Imperial
Calexico
VI11-1-32
NMNH
t
Chlosyne lacinia
USA California
Imperial
Calexico
June 1940
NMNH
5
Chlosyne lacinia
USA California
Imperial
El Centro
June 1940
Janes H Baker
NMNH
3
Chlosyne lacinia
USA California
Imperial
El Centro
June 1940
NMNH
1
Chlosyne lacinia
A USA California
Impend
El Centro, dead in store window
20 July 1957
TC Emmel
FSCA
Chlosyne lacinia
USA California
Imperial
Hdtville
7-23-35
E C J
NMNH
1
Chlosyne lacinia
USA California
Holtville
7-23-35
SE CN
NMNH
5
Chlosyne lacinia
USA California Kings
Armona
7-23-35
E C J Ernest Shoemaker
NMNH
Chlosyne lacinia
USA California Riverside Coachella
X-5-41
NMNH
Chlosyne lacinia
USA California Riverside Indio
10 May 1932 CM Danmers NMNH
I 2
Chlosyne lacinia
USA California Rivers.de Indio
5 Sep 1932
J W Tilden NMNH
.
Chlosyne lacinia
USA California Riverside Indio
13 Sept 1933
C M Danmers NMNH
:
Chlosyne lacinia
USA California Rivers.de Indio
4 Oct 1933
C M Danmers NMNH
Chlosyne lacinia
L @USA California @SanDego
@San Miguel Mts
VI-14-59
.JC Hoping
NMNH
Chlosyne lacinia
USA California ?
Necea
7 April 1929
Tom Abor
NMNH
Chlosyne lacinia
USA California [7
28 May 1932
C M Danmers
NMNH
2
Chlosyne lacinia
USA California 7
28 May 1932
C M Danmers
NMNH
,
Chlosyne lacmia
A USA Kansas Riley
2 July 1968
Gay Hevel
NMNH
Note NMNH New Mexico records are the croc ale phenotype, including specimens with narrow pale white cream bands and a few with narrow yellow-yellow/orange bands (beginnings of the transition to the ad|utnx phenotype
Chlosyne lacinia
c/a USA
New Mexico Donna Ana
Hatch
11-IX-1981
H L King
FSCA
Chlosyne lacinia
c/a USA
New Mexico
Donna Ana
Las Cruces ,8-21-31
FSCA
1
Chlosyne lacinia
A
USA
New Mexico
Eddy
Hope, 4500' 12-IX-1981
H L King
FSCA
2
Chlosyne lacinia
c
USA
New Mexico
Grant
Pinos Altos Mts. 4 July 1958
J P & G C Hubbad
NMNH
2
Chlosyne lacmia
c
USA
New Mexico
Grant
Pinos Altos Mts 26 July 1958
J P Hubbad
NMNH
1
Chlosyne lacinia
C/A
USA
(New Mexico?)
|Hddgo?|
5 miles NE Rodeo 30.VIII.58
H V Weems Jr
FSCA
1
Chlosyne lacinia
C/A
USA
New Mexico
Hidalgo
Rodeo, on desert 14 IX 58
HV Weems Jr
FSCA
1
Chlosyne lacinia
C
USA
New Mexico
7
Water Canon (=canyon?|, 5000 ft Aug. VI
F H Snow
NMNH
3
Chlosyne lacinia
C
USA
New Mexico
7
[illegible), on sunflower 1 Sep 98
NMNH
1
Chlosyne lacinia
C
USA
New Mexico
7
F H Snow
NMNH
1
Chlosyne lacima
0
USA
New Mexico
7
F H Snow
NMNH
+
Chlosyne lacinia
A
USA
Nevada
Clark
Glendale 2-IX-198I
H L King
FSCA
1
Chlosyne lacinia
A/C
USA
Nevada
Clark
Glendale 2-IX-1981
HLKrng
FSCA
7
Chlosyne lacinia
AC
USA
Nevada
Clark Glendale 20-IX-1961
HLfOn,
FSCA
3
Chlosyne lacinia
C
USA
Nevada
Clark 2-IX-1981
H L King
FSCA
2
Chlosyne lacinia
A
USA
Oklahoma
Cleveland Norman 1710 1950
W J Ronthai
FSCA
1
Chlosyne lacinia
A.
USA Oklahoma
Cleveland Norman 18 10 1950
W J Ran that
FSCA
2
Chlosyne lacmia
A
USA Oklahoma
Cleveland Norman 21 10.1950
W J Ronthai
FSCA
7
Chlosyne lacmia
A
USA Oklahoma
Cleveland Norman 122.10.1950
W J Ronthai
FSCA
2
Chlosyne lacinia
A
USA Oklahoma
Cleveland
Norman 26 10 1950
W J Ronthai
FSCA
1
Chlosyne lacinia
A
USA Texas
Bee
Beeville Aug 20 11
NMNH
1
Chlosyne lacinia
A
USA Texas
Bee
Beeville 19-1-16
FSCA
1
Chlosyne lacinia
A
USA Texas
Bee
Beeville 9-11-16
FSCA
1
Chlosyne lacmia
A
USA Texas
Bee
Beeville ¡IX-10-1916
Ernest Shoemaker
NMNH
1
Chlosyne lacinia
A
USA Texas
Bee
Beeville IX-20-1916
Ernest Shoemaker
NMNH
1
Chlosyne lacinia
A
USA
Texas
Bee
Beeville ¡IX-26-1916
NMNH
1
Chlosyne lacinia
A USA
Texas
Bee
Beeville Sept 23 16
NMNH
,
Chlosyne lacinia
A USA
Texas
Bee Beeville 2 Oct 1916
NMNH
,
Chlosyne lacmia
A/c USA Texas
Bee Beeville Oct 1920
John Adams Comstock
NMNH
2
Chlosyne lacinia
a USA Texas
Bee Beeville Aug 21-16
NMNH
,
Chlosyne lacinia
a USA Texas
Bee Beeville Oct 21-16
NMNH
Chlosyne lacinia
a USA Texas
Bee Beeville 15.V65
HOhWlon
FSCA
,
Chlosyne lacinia
a USA Texas
Bexar Fort Sam Houston 12 VII 1949
Jablonski
NMNH
Chlosyne lacima
a USA Texas
Bexar San Antonio 28 July 1919
GW Raw son
NMNH
Chlosyne lacmia
a USA Texas
Bex
San Antonio 29 July 1919
GW Rawson
NMNH
Chlosyne lacmia
a USA Texas
Bexar
San Antonio July 30/19
GW Rawson
NMNH
,
Chlosyne lacinia
a USA
Texas
Bexar
San Antonio 21 Aug 1919
JW Rawson
NMNH
Chlosyne lacinia
A
USA
Texas
Bex
San Antonio, attacking Helianlhis July 20 21
C H Gable (same ei below)
NMNH
Chlosyne lacinia
A
USA
Texas
Bex
San Antonio 24 July 1921
C H Gabie
NMNH
Chlosyne lacinia
A
USA
Texas
Bex
San Antonio May 1937
J P & G C Hubbad
NMNH
,
Chlosyne lacinia
A
USA
Texas
Bex
San Antonio, Olmas Pk X-26-1939
B Struck
NMNH
Chlosyne lacinia
A
USA
Texas
Bex
San Antonio, Olmas Pk | Vll-I-1940
NMNH
Chlosyne lacinia
A
USA
Texas
Bex
San Antonio ¡6-7-1947
Joe M Good Jr.
FSCA
3
Chlosyne lacinia
A/C
USA
Texas
Bex
San Antonio June 1957
J P 4 G C Hubbad
NMNH
Chlosyne lacinia
A
USA
Texas
Bex
San Antonio 20 July 1957
GW Rawson
NMNH
Chlosyne lacinia
c
USA Texas
Bex
San Antonw 20 July 1958
J P 6 G C Hubbad
NMNH
Chlosyne lacima
A
USA Texas
3exar
San Antonio 29 Mach 1963
John F Ronert
NMNH
Chlosyne lacima
A
USA Texas
Bex
4-VIII-72
FSCA
Chlosyne lacima
A
USA Texas
Bex
San Antonio 20 July 1979
GW Rawson
NMNH
,
Chlosyne lacima
A
USA Texas
Bex San Antomo VI-21-1979
G P Enge*idt
NMNH
Chlosyne lacima
a USA Texas
Bex San Antonio 18 Aug 1979
L Finkelston
,
Chlosyne lacmia
A
USA Texas
Bex San Antomo March 8-15
NMNH
Chlosyne lacima
a
USA Texas
Bex San Antonio April 1-7
NMNH
Chlosyne lacima
A
USA Texas
Bex San Antonio April 1-7
EdwTOwen
NMNH
,
Chlosyne lacmia
A
USA Texas
Bex San Antonio April 1-7
Banes
NMNH
Chlosyne lacinia
A
USA
Texas
Bex San Antonio May 1-7
NMNH
Chlosyne lacinia
A
USA
Texas
Bexa San Antonio
NMNH
4
Chlosyne lacmia
A
USA
Texas
Blanco Austin ¡6/21/73
E C Olson
Chlosyne lacinia
A
USA
Texas
Blanco
Johnson ¡20-V-?
Chlosyne lacinia
A
USA
Texas
Brewster
Black Camp 29-V-31
FSCA
,
Chlosyne lacinia
A
USA Texas
Cameron
Brownsville June 07
W A Twelkemeier
NMNH
Chlosyne lacinia
A
USA Texas
Cameron
Brownsville June 07
NMNH
Chlosyne lacinia
A
USA Texas
Cameron
3rownsville X 24.28
Chlosyne lacmia
a USA Texas
Cameron
Brownsville 13 VIII 1933
NMNH
,
Chlosyne lacinia
A USA Texas Cameron Brownsville 20X 1933
NMNH
,
Chlosyne lacinia
A USA ¡Texas Cameron Brownsville 21 X 1933
NMNH
Chlosyne lacinia
A USA Texas Cameron Brownswke 21 Mar'S9
W 4hm
NMNH
2
Chlosyne lacima
A U$A Texas Cameron Brownsville i April 1961
R He.t;m an
FSCA
1
Chlosyne lacinia
* USA Texas Cameron Brownsville 26 Jime 1966
Chlosyne ladnia
Chlosyne lacinia j
Chlosyne ladnia
A USA Texas Cameron near Brownsville 26 June 1968
FSCA
1
A
USA Texas
:ameron
rownsville 6-IV-75 i
NMNH
2
Chlosyne lacinia
A
USA Texas
3 ameren
rownsville 1 Sept 1979
L Finkelston
Chlosyne lacinia
A
USA Texas
3 ana on
rownsville 7-tl ¡
Chlosyne lacinia
A
USA Texas
:ameron
rownsville 10-22
Chlosyne lacinia
A
USA Texas
:neron
rownsville 10-24
A
JSA Texas
3 am or on
3


122
two plates wide enough that the ventral genital opening can be clearly seen in ventral
posterior view without prying the plates apart with a forceps.
Further description: Males: Valvae with ventral valve projections either entire or with
one small additional distal tooth. Ventral valve projections with a convex anterior side
and concave posterior side. Posterior part of valve with hair like setae containing a
distinct, hollow, posterior pointed projection except in the most primitive genus (Antillea)
which lacks a posterior projection. The inner valve process is entire and sharply pointed
or bluntly rounded off distally, and while always orientated posteriorly, it is variable
within the subtribe with respect to the posterior and anterior sides being convex or
concave. Valvae do not taper posteriorly, excluding distinct posterior projections.
Sockets for setae on inner sides of valvae visible or not in ventral view. A patch of short
setae is variably present or absent on the dorsal half of the posterior lateral sides of the
valvae anterior to the area with dense hair like posterior setae. In lateral view valvae
orientated primarily anterior to posterior, widely separated in ventral view. The ventral
valvae have an opening or slit located closest to the inner side of the valve, never
extending anterior to the vinculum. Surface ofjuxta variable but never smooth, always
with a ridge or plateau. Juxta lacking projections. Saccus forked (deeply or slightly)
with a pair of invaginated projections with their openings concealed by the vinculum or
exposed. Tegumen reduced to a simple bridge between the valvae, lacking posterior
projections. Ventral posterior side of phallus sclerotization pattern varies among genera,
and varies with respect to whether the lightly sclerotized area extends posterior or
anterior of the supersensory membrane. Dorsal posterior side of phallus uniformly


782
PLATE O
'////i


120
Higginsius Hemming
Synapomorphies from binary characters: Ventral valve opening slit-shaped (Figs. 87 &
89) (this is a homoplastic state which appears to have independently evolved in Antillea).
Inner valve process only slightly curved with the sides neither distinctly concave or
convex (Figs. 86 & 89).
Terminal derived states of multistate characters: Saccus tapering anteriorly and
triangular, then forming a narrow extension only slightly forked at the anterior end (Figs.
87 & 89). The quadrate shape of the plate on the ventral corpus bursae (Fig. 276) may
provide an additional character, however I was only able to examine females of one of
the two species (H. miriam).
Further description: Males: Inner valve process entire. Ventral valve process also entire.
Posterior edge of valvae with a hollow posterior projection in one species (H. fasciatus)
and without projections in the other (H. miriam). The shape and orientation of the
posterior process in H. fasciatus is unlike that of any of the projections found in the
Chlosyniti, and phylogenetic evidence indicates its presence represents an independently
acquired homoplastic state with respect to the Chlosyniti. Ventral valve opening
terminates posterior to the vinculum. Surface of juxta smooth, lacking ridges or plateaus.
Tegumen projections much shorter than the length of the tegumen anterior to the
projections.
Females (Based on the examination of H. miriam but not H. fasciatus)-. Ventral plate on
corpus bursae lacks paired anterior extensions or inverted teeth. Inverted teeth on corpus
bursae concolorous with corpus bursae and arranged as in Atlantea.


316
123. A dark red to red-orange ventral hindwing median band composed of a series of
patches distinct against the background color as described above:
0=Absent.
l=Present.
2=A light orange band is present.
This band also appears on the dorsal wing surface in the equivalent position for some
taxa, but its presence/absence on the dorsal surface is variable within some taxa, in
contrast to the ventral surface.
Considerable independent evidence from other characters suggests the band in C.
janais, C. marianna, and C. rosita represents an independent acquisition compared with
the band present in C. hippodrome. For characters 125 and 126 I elect to code taxa
lacking character 123 state 1 as 0 and C. hippodrome as although C. hippodrome
could be assigned state 2 for both characters. This action would represent weighting a
almost certain homoplastic acquisition three times instead of once. If C. hippodrome
were coded state 2 and taxa lacking character 123 state 1 were coded after the
computer fdls in a state for the ? taxa, the state of C. hippodrome would contribute to
the polarization of the C.janais/C. rosita ancestral node, which is not appropriate if the
character was independently acquired. For character 124, C. hippodrome has a unique
state and can be coded as such without any undue bias.
124. For those taxa with state 1 of the preceding character, the cells occupied by the
ventral hindwing median band:
*l=Prominent patches in CuA2-R5 (sometimes with a small diffuse patch in
Sc+Rl).


15
to warrant subspecies recognition, once again taxa are being delimitated by arbitrary
opinions as opposed to testable hypotheses.


429
Strict
Equal char, weights:
0=930
Rl=.970
RC=.903
Successive char, weights:
CI=.967
RI=,986
RC=.953
Boot strap scores
are based on
successively
weighted characters
95
62
87
78
94
91
52
87
55
"52
73
65
52
91
88
75
99
- C. nycteis
- C. gorgone
- C. harrissii
- C. kendallorum
- C. hoffmanni
- C. palla
- C. gabbii
- C. acastus
- C. w. damoetas
- C. w. whitneyi
- C. definita
- T. ezra
T. theona
- T. perlula
T. chinatiensis
T. I. leanira
T. I. fulvia
T. cyneas
T. cynisca
C. e. pardelina
C. melitaeoides
C. marina
C. e. poecile
C. e. erodyle
C. melanarge
C. eumeda
C. cal ¡fornica
C. lacinia
C. ehrenbergi
C. hippodrome
C. narva
C. g. gaudealis
C. janais
C. marianna
C. rosita rosita
C. montana
C. rosita browni
C. riobalensis
T. eleda
T. coracara
T. anomalus
D. dymas
M. elva
A. pelops
A. proclea
P. minuta
H. fasciatus
H. miriam
A. tulita
A. pantoni
CumPhyGnaMel
Figure 333: Strict consensus tree from a heuristic search of equally or successively weighted genitalic
characters from the Chlosyniti/Poladryiti data matrix with all pattern characters (char 53-144) deleted.


301
most C. chinatiensis. C. theona, C. chinatiensis and C. ezra are coded as polymorphic
for states 1 and 2.
In males of the out group taxon H. miriam, the postmedian dots are present but
very reduced, and sometimes composed of only a few cream colored scales. I code H.
fasciatus as because due to the extensive mottling of the wing and reduction of the
postmedian dots apparent in H. miriam, I was unable to determine if H. fasciatus has a
postmedian row of dots.
Antillea proclea always has one sharp white apical postmedian dot. Antillea
pelops usually has no postmedian dots, but a couple individuals were found with a pale
orange dot in the same position as the sharp white dot of A. proclea.
108. For those taxa with a forewing postmedian row of dots, the color of these dots.
0=Cream-colored (except dots in the R cells may be whiter).
1=A11 dots a clean, pure white.
#2=Light orange (see below)
Females of the out group taxon Atlantea pantoni have deep yellow postmedian dots while
males have light orange postmedian dots, while both sexes of Atlantea tulita have light
orange postmedian dots. I code Atlantea pantoni as state 2; since the females having
deep yellow postmedian dots is unique to this taxon there is no point in splitting this
feature into two characters for males and females. Furthermore, coding A. pantoni with a
separate sexually dimorphic state would lose the information that one sex shares the
unusual feature of light orange postmedian dots with another taxon.


275
*l=Cream colored with a variable mix of brown scales (brown scales may be
entirely absent).
*2=Composed of all black scales.
#3=A mosaic of black, orange, and white scales.
4=A intraspecifically variable mosaic of black and orange scales, with the black
scales tending to dominate.
Not all of the variation between this and the preceding character is likely to be
independent, although there are some taxa coded differently for these characters. This
potential problem is irrelevant for state 1 and state 2, as these represent autapomorphies.
85. On the forewing underside, presence of a distinct orange patch in the basal most part
of cell C, distinct against the background. If this feature is absent, the characteristic of
the basal part of cell C is described.
0=Absent, white scales in this area.
l=Absent, a mix of orange and white scales.
2=Absent, orange on the upper and lower sides of the cell with predominantly
white in the middle.
3=Absent, the area contains black scales.
4=Present.
5=Absent, the area is orange and continuous with surrounding orange but there is
no distinct patch.
6=Absent, but a distinct more elongate cream-colored patch is present (this patch
may be bordered costally by a thin row of dark orange scales.


367
pattern characters have all three tree statistics equal to 1.0, while some individual
genitalic characters have tree statistics lower than most individual pattern characters.
Variation in Boot Strap Scores with Different Numbers of Taxa in the Analysis
The boot strap technique has widespread use in systematics (Siebert 1992) and
involves randomly sampling characters until a data set the same size as the original data
set is obtained (Felsenstein 1985). A most parsimonious tree is calculated (or
approximated by heuristic algorithms in the case of large data sets) from each boot strap
data set and the boot strap score for a particular clade reflects the percentage of the boot
strap data sets which yielded that clade. While this technique does not yield what
statisticians would consider true confidence limits (statisticians have not yet been able to
calculate actual confidence limits for tree topologies), the percentage scores for particular
clades have been regarded as an index of their support (Siebert 1992). I argue that the
Chlosyniti/Poladryiti data set provides a useful case study for testing the hypotheses that
boot strap scores provide an indication of support for particular clades, since (1) the
Chlosyniti/Poladryiti data matrix yields a well resolved phylogenetic hypothesis
including a number of monophyletic groups of various sizes (2) this hypothesis varies
little as a result of whether characters are equally or successively weighted or how
polymorphic characters are analyzed, (3) the phylogeny is based on well grounded
hypotheses of character polarization, and (4) all known extant in group taxa are included
in the analysis (minus A. cryptadia and C. hylaeus, thae latter which may actually be
merely a form of C. eumeda).


321
2=No distinct crescents, but the color is solid between the postmedian band to the
submarginal band.
#3=A double row of crescents, composed of a generally narrow basal row and a
very diffuse and narrower distal row.
130. For taxa with states 1 and 2 of the preceding character, the color of these ventral
hindwing patches:
0=White.
l=Yellow (with a slight greenish tint).
*2=Pale yellow basally with small orange patches distally.
3=Yellow lacking any trace of a greenish tint.
#4=Pale orange.
@*A=0&3: Varies geographically along a continuum between state 0 and state 3.
@*#B=4&5: Sexually dimorphic, pale orange in the male, pale tan to light brown
in the female (see below for an explanation of this coding).
I chose not to code this character separately for males and females because the only
sexually dimorphic taxon is Higginsius miriam (although I did not have females of H.
fasciatus available) and state 5 (the female form) is autapomorphic. Coding H. miriam as
an unique autapomorphic state would lose the information that the males of the two
Higginsius species share a unique character state.
131. Dorsal hindwing upperside with a third well defined band of light colored patches(
bordered basally and distally by black and w. black along veins), located between the
basal median line and the distal band of the central symmetry system:
0=Absent.


123
sclerotized. Distal teeth bearing plate on vesica with some teeth distinctly larger than
others. Vesica otherwise simple, lacking teeth disjunct from plates or granulse patches.
Females: Ventral plate on corpus bursae with inverted teeth and paired posterior
extensions. Longitudinal extent of the corpus bursae free of the ventral plate is variable,
and often less than half of the corpus bursae length. Inverted teeth on the corpus bursae
distinctly sclerotized or concolorous with the corpus bursae, and apparently were
independently lost twice within the subtribe. When present, the inverted teeth are
arranged in two disjunct patches on the lateral sides of the corpus bursae. Ductus bursae
greatly reduced and membranous. Ostium bursae absent. Lamella postvaginallis and
antevaginallis well developed, and lamella antevaginallis variable for the presence or
absence of a distinct ridge. Lamella postvaginallis with a lightly sclerotized area around
the genital opening.
Remarks: The Chlosyniti are the subject of a more detailed phylogenetic analysis
covered in the next chapter, and a detailed revision in Chapter 4 including descriptions of
the three genera which I recognize.
Distribution: This subtribe is exclusively New World, occurring throughout North
America (except in the far north), Central America, and with a few representatives
ranging into northern South America. Representatives also occur in the Caribbean,
including one endemic genus comprising two species.
Key to the Subtribes of the Melitaeini Based on Genitalic Characters
1. Males: Inner projection of valvae with a posterior fork (slight in Eurodryas) (Figs.6-
14), inner wall of valvae with five or more thickened spine-like setae (Figs. 2-3 & 5-6),
medial area of saccus (between paired saccular projections) reduced to an extremely


538
specimens I have examined of malcom than most (but not all) specimens I have
examined of palla, and with regard to the former most evident with the aid of a dissecting
microscope particularly in specimens not in fresh condition. One statement from Behr's
Latin description (translated in J. Emmel et al. 1998c) does not in my view apply well to
malcom, palla, or any similar taxon but is more consistent with palla: "The wings of the
male close to a true red above". All specimens I have examined of malcom are pale
orange, while some specimens of palla (and gabbii and acastus) have reddish orange
upperside markings although they do not dominate. However, those palla with red
orange markings also have a more marked contrast between the median and postmedian
bands.
J. Emmel et al. (1998c) concluded "the association of this name [whitneyi] with
the high elevation Sierran C. palla...is erroneous". Given the absence of any specimens
or figures which can be associated with Behr's original specimens and the problematic
"wings of the male close to a true red" statement in Behr's (1863) description I do not feel
quite this strong of a claim is warranted. However, I do feel consideration of the
available evidence (the absence of any C. palla specimens known from the vicinity of the
clearly established type locality of whitneyi, the presence of malcom in this area, and
Behr's description overall fitting more closely with malcom), favors the hypothesis that
whitneyi and malcom are the same entity over the hypothesis that whitneyi is a synonym
of palla. This hypothesis is not undermined by considerations of nomenclatural stability
because both alternatives appear in major works on North American butterflies.
Consequently, I choose to adopt the views of Scott (1986) and J. Emmel et al. (1998c)
and regard malcom Comstock as a synonym of whitneyi Behr.


670
and C. hippodrome. The separation of C. lacinia from C. californica is provided in the
diagnosis of C. californica (above). All of the other Chlosyne species mentioned may be
separated from any form of C. lacinia by the absence of an orange patch in ventral
hindwing cell CuA2 that reaches nearly as far distally or surpasses the white postmedian
dot in that cell. Additional characters for separation are provided in the diagnosis
sections for the other taxa. The genitalia are identical to C. californica (above).
Geographic Variation: Chlosyne lacinia is the most intraspecifically variable of all
Chlosyne, and no less than twenty-three names are available for this one species. Yet, I
find the entire range of variation occurs along a continuum, with the exception of the
pattern of the meso- and metathoracic femur which can be polymorphic among
individuals from the same locality, and consequently I recognize no subspecies. There is
one odd specimen in the NMNH from the collection of the Brooklyn Museum with no
associated data that does not fall within the continuum of variation and probably
represents an aberration. The continuum of variation is not linear, but rather a complex
branching continuum with reconnections and dead ends, as depicted in Figure 345.
For the sole purpose of describing the wide range and complex patterns of
geographic variation, I find it convenient to retain the names of several forms
representing points along the continuum of variation, including adjutrix, crocale,
quehtala, lacinia, saundersi, and paupera. However, I stress these forms are not distinct
and do not represent subspecies or taxa of any rank, and there are no characters which
separate them because every form integrades with one to three additional forms. A
variety of names are available for intermediates between these forms as well as additional
names for individuals within the smaller rage of variation occurring within populations


798
Natural History-Smithsonian Institution, where he was first introduced to the methods of
phylogenetic analysis. Here, he worked on a project involving a phylogenetic study of
the genus Satyrium (Lepidoptera: Lycaeinidae) with Smithsonian research scientist
Robert K. Robbins, and also benefited greatly from the advice and assistance of Donald J.
Harvey. In 1996 he moved to Florida to begin work on a PhD in Lepidoptera systematics
under the supervision of Thomas C. Emmel. After considering a variety of potential
groups to study, he chose the genus Chlosyne and its relatives, in part due to the complex
morphological variation present in this group and the extensive amount of material
available for study. Soon afterwards, he decided that to work out a detailed phylogeny of
the Chlosyne, the systematics of the Melitaeini would need examination first, and he
expanded the project into this area as well. Phylogenetic studies of the Melitaeini and the
Chlosyniti became the focus of his graduate research, which proved fruitful in working
out a well-resolved phylogenetic hypothesis and natural classification for this group of
butterflies. In addition, as a graduate student he was employed as a teaching assistant in
the University of Florida's Department of Entomology and Nematology for five
semesters, where he assisted with courses on Insect Classification, Principles of
Entomology, Behavioral Ecology and Systematics, and Entomological Techniques.
These experiences provided him the opportunity to share his knowledge and experience
regarding insect systematics, classification, identification, evolution, collecting, and
curating with a number of dedicated students. Subsequent to graduation, he hopes to
continue to find the opportunity to pursue research interests in the systematics and
biodiversity of Lepidoptera.
797


614
are between the central symmetry system bands in cells CuA2-CuAl, the area between
the hindwing median and postmedian bands in cells CuAl to CuA2, and last of all the
hindwing basal area. The transition to the alma phenotype involves a change in forewing
shape, such that both sexes have more elongate forewings and that males have narrower
forewings with a more pronounced apex relative to females.
Range: I have examined specimens of C. leanira leanira from California, southern
Oregon, Nevada, Utah, Colorado ("black ridge"), and a single specimen labeled from
northeast Oklahoma (15 miles NW Tulsa, river bottom, May 30 1938, G. W. Rawson).
Data for specimens in the NMNH and FSCA are presented in Table 12. The distribution
maps of Austin and Smith (1998b) and Stanford and Opler (1993) include many specific
locality and county distributional records, respectively, from these areas (excluding
Oklahoma) with Colorado records limited to the eastern edge of that state. Brown et al.
(1992) (for T. leanira wrighti), Scott (1986) (for C. leanira leanira+C. leanira alma),
Stanford and Opler (1993) (for T. leanira (incl alma)), and Austin and Smith (1998b) (for
T. leanira) all report records from the Mexican state of Baja California Norte, and the
latter three sources also include the northwest edge of Arizona in the range. The
Oklahoma specimen is well east beyond the range included based on the combination of
specimens I have examined and records reported in the literature. It is particularly
intriguing because it is a specimen of the most orange alma phenotype of C. leanira
leanira well inside the range of C. leanira fulvia, but I am not willing go as far as to
consider this evidence of sympatric occurrence between leanira and fulvia based on a
single specimen, unless corroborating evidence would be obtained. For now, I suspect


471
ventral view. Ventral valvae with a distinct opening. Setae sockets on the inner valvae
walls in the vicinity of the midline visible in ventral view. A patch of short setae on the
outer lateral posterior sides of the valvae absent except in the most primitive taxon,
Microtia elecla. Saccus forks prominent in some representatives but very slight in others.
Female Genitalia: Inverted teeth on corpus bursae present and nonsclerotized in some
representatives, while absent or at least not detectable at 75X magnification in others.
Edges of lamellae antevaginallis uniformly sclerotized relative to the middle. Pouch
formed around the ventral genital opening by the lamella postvaginallis and lamella
antevaginallis tends to be more open relative to many other Chlosyniti. Lightly
sclerotized area on the lamella postvaginallis around the genital opening noncontiguous
with the posterior edge of the plate except in Microtia elva.
Pattern: There is considerable divergence in pattern characters within the genus Microtia
such that few generalizations can be made for the genus. M. elva has a wing pattern so
divergent that hypotheses of homology between wing pattern elements in that taxa and
other Melitaeini can not be determined. M. eleda and M. dymas have a very similar
primitive wing pattern, while M. coracara and M. anomalus share many derived features.
Pattern of body and appendages: Black scales always present on the labial palpi. Tuft of
white hairs on the frontoclypeus anterior-lateral to the base of each antennae (may be
very reduced and even absent in M. elva). Ventral surface of the abdomen with two
longitudinal dark stripes against a light background. Orange scales absent on the
antennal shaft (except in some specimens of T. eleda).
Wing Pattern: A narrow band of orange scales along the basal part of the forewing costal
margin. Dorsal hindwing without a band of colored patches between the position of the


20
I examined a number of taxa within Nymphalini and Kallimini to try to avoid
making unwarranted assumptions on character polarization based on missing data. I
adopt a similar out group approach to that of Robbins (1991) for combining Nymphalini
and Kallimini as the out group for a phylogenetic study of the Melitaeini, with my coding
guidelines outlined below. Some of my coding decisions are different from those of
Robbins (1991) in order to provide more specific information in the matrix, such as
distinguishing among the three scenarios where Robbins coded the out group 9.
However, their affect on the phylogenetic analysis is unchanged. For the purposes of
discussion, I refer to this method as the Cumulative Out Group Method.
I designate c as the symbol for a state occurring in out group taxa that never
occurs within the in group, and n and p as different character states that occur in both
the in group and out group (in the actual data matrix n or p would be designated as a
numerical state, except for multistate characters with more than 10 states where state
assignments follow the sequence 0,l,2,3,4,5,6,7,8,9,a,b,d (letters are used because
MacClade 3.07 interprets 10 as 0 & 1). The following is a summary of how I coded
characters for the cumulative out group: 1) All out group taxa exhibit a state (n) which
also occurs in the in group: the out group is coded n. 2) Some out group taxa have
state n, while others have state p, and both states n and p occur within the in group: the
out group is coded n,p (=n&p in MacClade) (if no states occur within the in group
other than n and p, coding the out group as ? or c would be equivalent in terms of
affecting the computer analysis, but less informative regarding the distribution of
character states among out group taxa). The same coding methodology would apply to
characters with three or more states occurring in both the out group and in group. 3) No


365
analogous to treating 0, 0&1, and 1 as three states of an ordered quantitative character,
with twice as many steps needed to go from state 0 to state 1 as from state 0 to state 0&1.
Consequently, an additional assumption of order is inherent to the DPCWH model which
is not present in the CALEOS model (but only between states participating in a
polymorphism), and thus the differences demonstrated in Scenarios 2 and 3.
Variation in Tree and Character Statistics with Different Numbers of Taxa
Siebert (1992) noted that the consistency index (Cl) (minimum number of
changes on a tree/actual number of changes) has been negatively correlated with the
number of taxa and characters in an analysis, thereby reducing its value in terms of
comparing different phylogenetic trees with different numbers of taxa and/or characters.
I find that the same is true with respect to the number of taxa in the analysis for the
retention index and rescaled consistency index, both for the values of these statistics for
the tree (Figure 328) and their average value for characters (Figure 329-322). In my case
study (Figure 328) the retention index was less impacted by adding taxa to the analysis
than the consistency index, and the rescaled consistency index (CI*RI) naturally was the
most impacted by adding taxa to the analysis. Furthermore, in my case study the
correlation between these tree statistics and the number of taxa in the analysis does not
appear to be a linear relationship, as would be expected since some taxa in the data set
appear to be more divergent (have evolved at a faster rate) relative to others, and
consequently the contributions of different taxa to homoplasy are far from uniform. I
suspect the nature of the pattern of decrease in tree statistics associated with adding
additional taxa will be variable with different data sets as opposed to constituting a more
widely occurring pattern. Consequently, even if it is accepted that the comparative value


689
Higgins (1960) reports records for the eastern slopes of Hidalgo, and mentions a
specimen labeled "Texas" which I view with suspicion, given the absence of any records
from northern Mexico among the specimens I have examined. Ferris (1989) credits the
species to the United States, but gives no further locality information, and it is unclear
whether or not the source is different from of Higgins (1960).
Species Delimitation: The genitalia of C. ehrenbergi are unique, with clear gaps in
certain genitalic characters compared with its closest relatives. Also, C. ehrenbergi has
no sister species. The wing pattern of C. ehrenbergi is also unique and markedly
different from any other Chlosyne.
Clilosyne hippodrome (Geyer)
Figures 151,171,192,202,232 & 249.
Araschnia hippodrome Geyer, 1837. Zutrage. 5:17. Type Locality: Mexico.
-Chlosyne hippodrome form fabricii Higgins, 1960. Trans. R. Ent. Soc. Lond. 112:400-
401.
Diagnosis: In collections, C. hippodrome has been confused with C. melanarge, C.
junis {gloriosa phenotype), and C. lacinia (quehtala phenotype). In C. hippodrome the
forewing markings are clean white versus cream-yellow in C. melanarge. C. hippodrome
has no dorsal red-orange postmedian band patch on the hindwing, and ventrally has white
postmedian dots but no postmedian band, while C. melanarge has a red-orange
postmedian band and at most one faint cream postmedian dot in cell M3. The red median


502
specimens I have seen from any other state or province. Series from Quebec and New
Brunswick contain a mix of very light, medium, and dark individuals. The darkest
specimens I have examined are from Pennsylvania, and series from that state are on the
average darker than series from any other state or province with the possible exception of
West Virginia (I have seen few specimens from that state). However, specimens from
the same localities as the darkest individuals include medium orange individuals well
within the normal range of variation throughout most of the range. I have not seen the
extreme dark individuals from neighboring New York or Ohio. Series from other states
and provinces listed below contain a mix of light, medium, and dark individuals, but none
as dark as the darkest Pennsylvania specimens and none as light as the most extreme
Manitoba specimens. Some Wisconsin specimens in my collection (Forest and Jackson
Counties) approach the extreme light condition but are still not as light as the most
extreme Manitoba specimens. The name liggetti has been used in association with the
dark phenotype while the name hanhami has been used in association with the light
phenotype (Bauer 1975).
Range: I have seen specimens of C. harrisii from Canada: New Brunswick, Nova Scotia,
Quebec, Ontario, and Manitoba. I have seen none north of New Brunswick; however,
Scott's (1986) range map includes Newfoundland. Stanford and Opler (1993) report a
record from southeast Saskatchewan. In the United States, I have seen specimens from
Maine, New Hampshire, Vermont, Massachusetts, Connecticut, New York,
Pennsylvania, New Jersey, Maryland, West Virginia (Randolph County, Spruce Vinab
Lk. (the most southern record that I have examined)), northern Ohio, Michigan, northern
Indiana, northern Illinois, northern and central Wisconsin, and Minnesota, and North


663
orange scales dorsally and throughout the distal end and white scales ventrally, tibia and
tarsi covered with orange scales only. Meso- and metapleurons where the femur tucks in
with black scales and hairs, with white scales sometimes visible underneath the dark
ones. Thorax ventral to the wings with a variable mix of white and black scales and
hairs. Antennal shaft with a checkered pattern of black and white scales without orange
scales, club with a white patch on the outer lateral side.
Forewing basal costal margin with a narrow band of orange scales. Wing fringes
checkered black and white with white sections between where the veins reach the wing.
The white section across from forewing cell Ml tends to be smaller than for the other
cells.
Dorsal forewing predominately black in basal area distal to the median band with
a variable amount of diffuse orange scaling over the black background. A light patch is
present in the forewing discal cell basal to the position of the discal spot, with an orange
border and cream scaling in the center. A smaller orange patch (usually with some cream
scales in the center) is also present in the dorsal forewing discal cell basal to the position
of the distal band of the basal symmetry system. The median band is broad and occupies
cells 1A+2A-R3. The band includes cream scaling basally and orange scaling distally,
with the orange area broader except in cells R3 and R4 and the cream scaling sometimes
absent in cell M3. Thin black scaling along the veins partially or completely separates
sections of the median band. Distal to the median band the dorsal forewing is black
except for postmedian dots and orange patches between the postmedian dots and wing
margin. The postmedian dots are sharp and white and present in cells CuA2-R5, R3, and
sometimes R4. The CuA2 dot is sometimes divided into two by the veinlet, and the R3


262
#4=Three distinct white patches at the posterior end of abdominal segments, a
dorsal patch and two bilaterally symmetrical lateral patches.
Some specimens of Chlosyne harrissii, C. nycteis, and C. gorgone have the posterior
white bands reduced and sometimes limited to light gray scales. Many specimens are
normal state 0, and I assign this state to these taxa, despite their exhibition of a greater
range of variation than other state 0 taxa.
A few taxa have no apparent bands but scattered white scales around the posterior edge
of their posterior abdominal segments but lack even broken light colored bands, including
C. gaudealis, C. marianna, and the gloriosa phenotype of C.janais. This pattern is
essentially an intermediate condition between state 0 and state 1. Since the other taxa
could be unequivocally scored 0 or 1 (or one of the other states) I code these taxa as "?".
I do not code them "0&1" because they do not exhibit either state but rather an
intermediate condition.
70. Distinct, solid, lateral orange stripes on the abdomen:
0=Absent.
* 1 =Present.
71. Presence of orange scales on the abdomen. There is also quantitative variation in the
amount of orange scales on the abdomen among taxa which was found to be too variable
intraspecifically to code.
0=Present.
1 =Absent.
@A=0&1: Individuals vary along a continuum between state 0 and state 1.
72. Pattern on the ventral surface of the abdomen:


526
postmedian band patches in cells CuA2-Ml and composed of cream colored scales with
thin black borders often not encircling the entire dots, with the dots tending to be more
conspicuous and black border often more extensive ventrally. The cell M3 dorsal
hindwing postmedian dot is dominated by cream colored scales with the black border
often limited to the distal side (sometimes absent), dots in cells CuAl and M2 like cell
M3 except the black border tends to be more extensive, dots in cells CuA2 and Ml may
still have cream scales dominating but tend to have more black than dots in other cells
with black occasionally dominating and rarely composing the entire dot. Distal and basal
median lines present on the ventral hindwing surface, distal sections straight or slightly
curved distally, basal sections straight or slightly curved basally, each section generally
lining up with those in adjacent cells. Dorsal hindwing and forewing margin between the
position of the marginal and submarginal bands black with a prominent clean orange
band (separated into sections by thin black scaling along the veins) extending from cells
CuA2 to Sc+Rl (occasionally just to cell R5). Ventral hindwing and forewing in the
same position with a sharp orange band. Ventral hindwing submarginal band distinct
against the background. Dorsal hindwing patches comprising the median band often
distinctly lighter than those comprising the forewing median band (light cream versus
darker cream, cream versus orange, or very light orange versus darker orange, with every
intermediate between these), but sometimes concolorous or only slightly different.
Dorsal hindwing between the submarginal band and distal parafocal element black often
with a well developed series of light colored patches with the basal side convex and the
distal side concave to straight; these patches are always prominent in the southern part of
the range and have a progressively increased tendency toward reduction moving


333
The arrangement of taxa within the Poladryiti, and of Antillea, Microtia,
Dymasia, Texola, and a clade with the other Chlosyniti is identical to that obtained from
the analyses of the Melitaeini data matrix in the preceding chapter (Figures 293-294 and
297-298), with Microtia and Dymasia coming out in the middle of Higgins (1981)
paraphyletic concept of Texola. Thessalia comes out as a monophyletic genus; however,
the results indicate it is one lineage within Higgins (1969 & 1981) paraphyletic concept
of the genus Chlosyne. Also, the monotypic genus Anemaca comes out in the middle of a
clade of derived Chlosyne. The validity of Charidryas was neither established nor
refuted, as two clades of Charidryas appear in a trichotomy with the clade containing the
rest of the Chlosyne (including Thessalia and Anemaca).
After two iterations of successive weighting the shortest trees were obtained,
including eleven equally parsimonious trees with a consistency index of 0.865, a
retention index of 0.937, and a rescaled consistency index of 0.811. The primary
difference between a strict consensus of these trees (Figure 312) and those generated
from the analysis with equally weighted characters (Figure 311) is that Charidryas comes
out as a monophyletic sister clade to the rest of the Chlosyne group. The only additional
difference is that trichotomies within the Thessalia leanira and Thessalia theona groups
were resolved. With 71 leanira group I refer to leanira, fulvia, cyneas, and cynisca,
and with 71 theona group I refer to theona, perlula, and chinatiensis.
Boot strap 50% consensus trees of equally and successively weighted characters
are presented in Figures 313 and 314, respectively. The boot strap 50% consensus tree
for successively weighted characters is better resolved than the one obtained for equally
weighted characters, but neither show as many clades as either parsimony analysis. No


544
"Mountains near Los Angeles", restricted to "La Tuna Canyon, Verdugo
Mountains, Los Angeles County, California, 1200' elevation" by J. Emmel et al
(1998c). Neotype: California Academy of Sciences (J. Emmel et al. 1998c.
figures 19-21).
-Melitaea sonorae Boisduval, 1869. Ann. Soc. ent. Belgique, 12:56. Type Locality:
"De la Sonora", restricted to La Tuna Canyon, 1200' elev., Verdugo Mts., Los
Angeles Co., California by J. Emmel et al. 1998d. Lectotype: National Museum
of Natural History (figures 246-248 in J. Emmel et al. 1998d).
=Melitaea pola Boisduval, 1869. Ann. Soc. ent. Belgique. Type Locality: "Sonora",
restricted to La Tuna Canyon, 1200' elev., Verdugo Mts., Los Angeles Co.,
California by J. Emmel et al. 1998d. Holotype: Carnegie Museum (figures 243-
245 in J. Emmel et al. 1998d).
=Chlosyne (Charidryas) gabbii atrifasciata Hawks and J. F. Emmel, 1998. Systematics
of Western North American Butterflies: 322-326 (figures 5-8). Type Locality:
California: Santa Barbara County, Santa Cruz Island, La Cascada. Holotype:
Natural History Museum of Los Angeles County.
Diagnosis: C. gabbii is extremely similar to some populations of C. acastus, but I have
found one character that appears to be a universally reliable species separation character.
In C. gabbii, the ventral hindwing background color is not uniform in contrast to C.
acastus. In both C. gabbii and C. acastus, the ventral hindwing white patches distal to
the postmedian band are pearly white (pale white in some acastus), however in C. gabbii
and not C. acastus there is a variable amount of yellow tinged cream scaling adjacent to


Table 12 Continued
738
Chlosyne lean a cyneas
Chlosyne lean a cyneas
Chlosyne lean a cyneas
Chlosyne lean a cyneas
Chlosyne leanir a cyneas
Chlosyne leana cyneas
Chlosyne lean a cyneas
Chlosyne leanira cyneas
Chlosyne leanira cyneas
Chlosyne leanira cyneas
Chlosyne lean a cyneas
USA
USA
USA
USA
USA
USA
USA
USA
USA
USA
USA
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Coch.se
Coch.se
Cochise
Coch.se
Cochise
Cochise
Cochise
Cochise
Cochise
Coch.se
Cochise
Huachuca Mis
Huachuca Mis
Huachuca Mts
Huachuca Mis
Huachuca Mts
Huachuca Mis
Huachuca Mis
Huachuca Mis
Huachuca Mis
Huachuca Mis
Chlosyne lean a cyneas
USA
Arizona
Cochise Ft Huachuca. lower Garden Canyon. 5.100"
10 Oct 1994
Douglas C Ferguson
NMNH
2
Chlosyne leanira cyneas
USA
Arizona
Cochise Palmertee
NMNH
7
Chlosyne leanira cyneas
USA
Arizona
Cochise Palmerlee
Barnes
NMNH
3
Chlosyne leanira cyneas
USA
Arizona
Cochise Palmerlee
EdwTOwen
NMNH
,
Chlosyne leanira cyneas
USA
Arizona
Cochise Paradise
NMNH
,
Chlosyne leanira cyneas
Mexico
Hidalgo
Incarnation
¡4-21-1972
H L King
FSCA
1
Chlosyne leanira cyneas
Mexico
Hidalgo
Rl 85. San Vincent
| 18.IV.63
R E Woodruff
FSCA
1
Chlosyne leanira cyneas
Mexico
Morelos
7000 fl
W M Schaus
NMNH
,
Chlosyne leanira cyneas
Mexico
Morelos
Popocatepetl Park
W Schaue
NMNH
,
Chlosyne leanira cyneas
Mexico
Morelos
Popocatepetl Park. 8-1000 ft
Juty
W Schaus
NMNH
Chlosyne leanira cyneas
Mexico
Morelos
Popocatepetl Park. 8-1000 ft
Chlosyne leanira cyneas
Mexico
Veracruz
Las Vigas
1 NMNH
1
Chlosyne leanira cyneas
Mexico
Veracruz
Las Vigas. C de Parte
W Schaus
NMNH
Chlosyne leanira cyneas
Mexico
?
6/19-36
WM
Chlosyne leana cyneas
Mexico
?
Hwy 190 SE. Mileage 62576.9
7-9-68
HV Weems Jr
¡FSCA
1
Chlosyne leanira cynlsca (NHM H.AM .[FSCA))
Chlosyne leanira cynisca
Mexico
Oaxaca
Car room a. Pan American
¡VI1-55
¡T Escalante
AM
,
Chlosyne leanira cynisca
Mexico
Oaxaca
San Jose Pacifico Mpio Rio-Hondo, 2400m
19X1971
!EC Welling
AM
,
Chlosyne leanira cynisca
Mexico
Oaxaca
San Jose Pacifico Mpio Rio-Hondo, 2400m
30X1971
EC Welling
AM
,
Chlosyne leanira cynisca
Mexico
Oaxaca
San Jose Pacifico Mpio Rio-Hondo. 2400m
24 XI 1971
ECWelhng
AM
,
Chlosyne leanira cynisca
Mexico
Oaxaca
Tam azul ap an
VI 28 56
T Escalante
AM
Chlosyne leanira cynisca
Mexico
Puebla
Tehuacan
Sept 10
R Muter
NMNH
t
Chlosyne leanira cynisca
Mexico
Puebla
Tehuacan
Oct 10
R Muller
NMNH
,
Chlosyne ezra (NMNH.FSCA)
Chlosyne ezra
Costa Rica
Limn
Guaprles
Schaus & Barnes
NMNH
Chlosyne ezra
Costa Rica
Limn
Port Limn
Jan-07
NMNH
Chlosyne ezra
Costa Rica
Limn
Port Limn
Jan "07
W M Schaus
NMNH
1
Chlosyne ezra
Costa Rica
Puntarenas
Osa Peninsula. Corcovado-Sieena
25 Jen 1977
NMNH
,
Chlosyne ezra
Costa Rica
Puntarenas
Llorona. Orsa Peninsula
21 Jan 1977
GBSmd
NMNH
,
Chlosyne ezra
Costa Rica
Puntarenas
Pan Ame. Hwy X San Vito Rd Ferry Crossing nr Biu|0 Rio Palmate General. 500
10/2/86
¡T C Emma!
FSCA
,
Chlosyne ezra
Costa Rica
Puntarenas
Patm Norte
9-18-70
FSCA
Chlosyne ezra
Costa Rica
Puntarenas
Palmar Norte
5-VI-1971
H L King
'FSCA
,
Chlospie ezra
Costa Rica
Puntarenas
Ptimar Norte
10-VI-1971
HLKny
'FSCA
Chlosyne ezra
Costa Rica
Puntarenas
Palmar Norte
115-VI-1972
H L King
Chlosyne ezra
Costa Rica
Puntarenas
Palmar Norte
16-VI-1972
HLKing
¡FSCA
6
Chlosyne ezra
Costa Rica
Puntarenas
Palmar Norte
26 VIII 1977
G B Small
NMNH
,
Chlosyne ezra
: Costa Rica
?
Es perange
Schaus
NMNH
Chlosyne ezra
Costa Rica
?
La Florida. 500 fl
May
Schaus & Barnes
NMNH
Chlosyne ezra
Panama
Code
Rio Hato
17 VII 1963
G B Small
NMNH
Chlosyne ezra
Panama
Chiriqui
NMNH
,
Chlosyne ezra
Panama
Chiriqui
Cerro La Galera de Chorcha
8-VIII-1975
G B Small
NMNH
Chlosyne ezra
Panama
Chiriqui
Rio Tole, 120m
17-VIII-73
NMNH
Chlosyne ezra
Panama
Darien
Serrana de Pitre Canee. tOOOrn
24-VII-1983
G B Smd
NMNH
~
Chlosyne ezra
Panama
Panama
Cerro Campana
29-1-1973
H L King
'FSCA
Chlosyne ezra
Panama
Canal Zone
Farfan C Z.
XI-11-62
NMNH
Chlosyne ezra
Panama
Old Panama
VIII-14-64
NMNH
Chlosyne ezra
Panama
NMNH
Chlosyne ezra
Panama
EdwTOwen
NMNH
Chlosyne ezra
Peru
B Neumogen
1
Chlosyne chinatiensis (NMNH.FSCA)
Chlosyne chinatiensis
USA Texas
Culberson 13 miles N of Van Horn
14 Sept 1950
j H V Daly
FSCA
1
Chlosyne chinatiensis
Chlosyne chinatiensis
USA Texas
USA Texas
Culberson 10 miles N of Van Horn, ex On Leucophytlum
Cutter son 12 mdee N of Van Horn
9Sapt 1969
26 Sept 1969
Jen Scott
FSCA
2
Chlosyne chnaUensis
Chlosyne chinatiensis
USA Texas
USA Texas
Culberson ¡12 miles north of Van Horn
Culberson 10 miles north of Van Horn
17 Sept 1970
25 Sept 1971
Jim Scon
FSCA
S
Chlosyne chinatiensis
Chlosyne chinatiensis
USA Texas
USA Texas
| Cubar son ] 10 miles west of Van Horn
Culberson 10 miles N of Van Horn, ex On Leucophyilum
125 Sapl 1971
26 June 1972
Jim Scott
John Haferntk
_FSCA
Chlosyne chinatiensis (F)
Mexico Duranqo
65 Mi N Durango, 4800'
19 Aug 1971
R K Robbins
I
June
July 16-23
Aug. 1-7
Aug. 8-15
Aug 16-23
Aug 24-30
Sep 1-7
Oct 8-15
VI-24-02
County (USA)
Locality (USA)
Taxon
Country
State/Prov.
Locality (non US)
Locality Continued (non USA)
Date
Collector
Collectior
No
Chlosyne leanira fulvia
USA
Colorado
La Plata
Durango
29 July 1972
RARahn
NMNH
2
Chlosyne leanira fulvia
USA
Colorado
La Plata
Junction Canyon
NMNH
1
Chlosyne leanira fuhna
USA
Colorado
Mesa
Black Ridge. 6800"
24 May 1971
Mike Fisher
FSCA
1
Chlosyne leanira fulvia
USA
Colorado
Pueblo
Burnt Mill Rd 11 miles S and W of Pueblo
R E Stanford
FSCA
1
Chlosyne leanira fulvia
USA
Colorado
San Juan
Silverlon
Aug. 8-15
Barnes
NMNH
4
Chlosyne leanira fulvia
USA
Colorado
June 16-23
Barnes
NMNH
,
Chlosyne leanira fulvia
USA
Colorado
NMNH 1
Chlosyne leanira fulvia
USA
Kansas
Barber
6 miles N of Hardtner
15 June 1971
BJ Smith
FSCA
6
Chlosyne leanira fulvia
USA
Kansas
Barber
8 miles W Med-cine Lodge short grass prairie along St Hwy 160
28 May 1974
J R Heitzman
FSCA
2
Chlosyne leanira fulvia
USA
Kansas
Barber
8 miles W Medicine Lodge shon grass prairie along St. Hwy 160
2 June 1974
J R Heitzman
FSCA
31
Chlosyne leanira fulvia
USA
Nebraska
Dawson
Hills along Tri-County Canal, larva. Castillera sesstliflora
6-5-87
SM Spooner
NMNH
2
Chlosyne lean a fulvia
USA
Nebraska
Dawson
Tri-County Canal
4-30-88
S M Spooner
NMNH
1
Chlosyne leanira fulvia
USA
Nevada
Lincoln
North Platte
9-1-88
S M Spooner
NMNH
1
Chlosyne leanira fulvia
USA
New Mexico
Grant
Chevey Creek, Pinos Altos Mts
25 June 1958
J P Hubbard
NMNH
,
Chlosyne leanira fulvia
USA
New Mexico
McKinley
Ft. Wingate
June 8-15
Barnes
NMNH
2
Chlosyne leanira fulvia
USA
New Mexico
McKinley
Ft Wingate
June 16-23
Banes
NMNH
7
Chlosyne leanira fulvia
USA
New Mexico
Scandoval
Jemez Springs
VI. 6
EdwTOwen
NMNH
2
Chlosyne leanira fulvia
USA
Oklahoma
Woodward
Alabaster Caverns St. Park near Freedom
11 May 1984
G F & J F Heve!
NMNH
1
Chlosyne leanira fulvia
USA
Texas
Culberson
Guadalupe Mts Natl. Park, Choza Spring. 5300 ft
20 June 1986
R Robbins
NMNH
e
Chlosyne leanira fulvia
USA
Texas
Terrel
10 miles W of Dryden
2-IX-75
FSCA
,
Chlosyne lean a fulvia
Texas
Terrel
10 miles W of Dryden
30-VIII-75
FSCA
t
Chlosyne leanira fulvia
USA
Texas i?
B Neumoegen
NMNH
,
Chlosyne leanira fulvia
USA
Utah
Kane
Buckskin Mt Area. 38 miles e of Kanab
20 May 1979
J Brock
NMNH
,
Chlosyne leanira fulvia
USA
Utah
Kane
Buckskin Mt Area. 38 miles e ol Kanab. 5000 It. Host: Castilleia chromosa. e larva 11 Apr 1980. adult 28 Apr. 1960:J Brock:
MNH
t
Chlosyne leanira fulvia
USA
Utah
Kane
The Cockscomb Hwy 89, 38 miles e of Kanab
J Brock
NMNH
i
Chlosyne leanira cyneas (NMNH.FSCA)
Chlosyne leanira cyneas
USA
Arizona
EdwTOwen
NMNH
i
Chlosyne leanira cyneas
USA
Arizona
So Arizona
Poling
NMNH
,
Chlosyne leanira cyneas
USA
Arizona
Cochise
Huachuca Mts
NMNH
,
Chlosyne leanira cyneas
USA
Arizona
Cochise
Huachuca Mts
July 24-30
EdwTOwen
NMNH
5
NMNH=Nat Mus, of Natural Hist.-Smithsonian Inst.; FSCA=FL St i
Sames 4 McDunnough
HAKaeber
Jim Brock
NMNH
NMNH
NMNH
NMNH
NMNH
NMNH


521
Minnesota, South Dakota, Kentucky, Missouri, Georgia, Alabama, Texas, Oklahoma,
Nebraska, Colorado, northeast Utah, and Montana. Scott's (1986) range map also
includes New York, Pennsylvania, southern West Virginia, Arkansas, North Dakota,
Idaho (a disjunct area in the southwest part of the state), Wyoming, and northeast New
Mexico. The extreme edges of the range include the North-West Territories: Areola
(Higgins 1960) (these records are far removed from any localities I have seen specimens
for or other reports in the literature), east central Alberta (Scott 1986), southwest Idaho
(Scott 1986, Stanford and Opler 1993), central Texas (Scott 1986, Stanford and Opler
1993, NMNH specimens) coastal South Carolina (Scott 1986, Gatrelle 1998), New York
(Scott 1986) White River, Ontario near the north shore of Lake Superior (NMNH
specimens), and eastern Ontario (Ross and Layberry 1998). Higgins (1960) mentions C.
gorgone occurs in Mexico but does not give a locality.
Species Delimitation: See C. nycteis (above).
Chlosyne hoffmanni (Behr)
Figures 138,162,180,209,217 & 238.
Melitaea hoffmanni Behr, 1863. Proc. Calif. Acad. Sci. 3:89. Type Locality:
"...Peculiar to the higher regions of California" (quoted from Behr (1863) in J.
Emmel et al. (1998c), restricted to "Gold Lake, Sierra County, California" by J.
Emmel et al. (1998c). Neotype: California Academy of Sciences, San Francisco
(J. Emmel et al. 1998c, figures 22-24).
-Melitaea hellicta Boisduval, 1869. Ann. Soc. ent. Belg. 12:55. Type Locality:


683
polymorphic within the same locality (California populations of C. whitneyi have orange
and white femurs while Colorado populations have orange and black femurs, but no
population exhibits both character states as in C. lacinia), as is the case where
crocale/quehtala, lacinia/sounder si, and quehtala/saundersi intermediates occur. I have
not seen specimens where the pattern on the femur exhibits an intermediate condition.
Meso- and metathoracic tibia and tarsi covered with orange scales only. Meso- and
metapleurons where the femur tucks in with black scales and hairs, with white scales
sometimes visible underneath the dark ones. Thorax ventral to the wings with a variable
mix of black and white scales and hairs. Antennal shaft with a checkered pattern of black
and white scales without orange scales, club with a white patch on the outer lateral side.
Forewing basal costal margin with a narrow band of orange scales. Wing fringes
checkered black and white with white sections between where the veins reach the wing.
The white section across from forewing cell Ml tends to be smaller than for the other
cells. Basal part of ventral forewing cell C solid black with a distinct patch of orange
which in some specimens is longer than in most taxa with this feature (as in C.
californica), but short in other individuals.
Hindwing postmedian band occupying cells CuA2-Sc+Rl and solid black (except
for the white postmedian dots) as in C. californica, sometimes distinct against a light
background but continuous with a dark background in other individuals (varying
geographically based on the forms described above). Postmedian dots white and
occupying cells CuA2-Sc+Rl on the hindwing, usually small or partially absent dorsally
and more conspicuous ventrally. The forewing postmedian dots are also white and tend
to be larger (they are never smaller) than on the hindwing, and occupy cells CuA2-R5


442
of the male genitalia, since these were studied in far greater detail on account of the
female genitalia providing no characters which could be coded into discrete states for
phylogenetic analysis within the genus Chlosyne. For taxa where the males could not be
separated genitalicly, I dissected larger series of females, but found no case where taxa
which could not be separated by male genitalia differed by female genitalia.
Species/Subspecies Further Description: This section includes the most detailed
description of each taxon. I have not written detailed accounts describing the genitalia
since they are illustrated in detail and genitalic differences between taxa and intrataxon
genitalic variability are discussed in the diagnosis sections. Characters of the body and
appendages are presented in roughly the same order that they appear in the character
descriptions for phylogenetic analysis. The order in which wing pattern characters are
described is not consistent between species taxa. The high variability in wing pattern
between different taxa, and the variability in the degree of similarity or difference
between dorsal and ventral patterns within the Chlosyniti, make attempting to follow a
consistent order impractical. Examples of the adults of each taxon are illustrated in
Plates A through V in Appendix A.
Variation/Geographic Variation: The range of variation exhibited by each taxon is
discussed, in addition to which characters vary geographically. With very few exceptions
for some of the rarest taxa, the series examined included substantial material from most
or all of the documented range of each taxon and therefore in my view are likely to
represent the full range of variation (excluding rare aberrations). However, there were
not always enough large series from many specific localities throughout the range of a
taxon to be able to describe localized patterns of differences in the average range or


197
Figures 279-282. Female genitalia of Microtia species in ventral aspect. 279:
M. coracara. 280: M. eleda. 281: M. dymas. 282: M. elva.
97(1),98(2)
Note: The apparent difference in
98(1) taxa is largely due to differences
in aspect. In Fig. 280 the LAV is more
ventral and less posterior relative to Fig.
282.
97(1)
98(2)
86(1)
(all Chlosyniti)
*
37.5X
282
280
1) (all
Chlosyniti)


372
solution of the entire data set cannot justifiably be claimed to be poorly supported for this
reason, or relatively poorly supported compared to clades which appear in a higher
proportion of such replications. At best, I argue a potential value of boot strap scores
may be that scores which are consistently high regardless of the number of taxa included
in the analysis could indicate well supported clades, since such clades repeatedly
represent a most parsimonious solution regardless of how data is randomly discarded or
disproportionately weighted. However, I personally reject the hypothesis that boot strap
scores provide relative indications of support for different clades, including the
hypothesis that boot strap analysis provides a evidence that clades with low boot strap
scores in a particular data set are more poorly supported by that data set than clades with
higher boot strap scores.
Equal Character Weighting, Successive Character Weighting, and Successive
Character State Weighting: An Alternative Method Proposed for Generating
Optimal Trees
Siebert (1992) suggested that character weighting, including successive
weighting, was one of four main ways for dealing with character conflict. He also
suggested that character weighting required additional assumptions that further remove
systematic analyses from their empirical basis and expressed the view there is good
reason for not doing it at all. I find this argument to be invalid, because every
phylogenetic analysis weights characters; successively weighting characters involves
character weighting just as much as equally weighting characters, only the two
approaches weight characters differently (i.e., equally weighting characters is not
synonymous with not weighting characters!). Likewise, I argue that successive weighting


126
272) OR distinctly greater than five times as long as the width of its anterior side (Fig.
267).
Distribution: Nearctic or Neotropical 4
3. Males: Entire valvae do not extend posteriorly and taper to a point (a posterior
pointed projection may or may not be present), a distinct, hollow ventral valve process
anterior to the part of the valve with extensive hair-like setae (Fig. 1).
Females: Ventral plate on the corpus bursae extending anteriorly as paired extensions OR
if not, the plate is not distinctly curved posteriorly AND no more than three times as long
as the width of its anterior side.
Distribution: Nearctic or Neotropical 5
4. Males: Valvae orientated much more posteriorly than ventrally (Figs. 60 & 64.1-
64.2), posterior inner wall of valvae extended inward in a flat, somewhat triangular
projection with the anterior side approximately perpendicular to the midline of the
genitalia and the posterior side slanted inward (Figs. 58-59 & 61-64), posterior end of
valve terminates in a elongate, tubular extension ending distally in a sharp, inward curved
point (Figs. 58-64.2).
Females: Sclerotized plate on the ventral corpus bursae not curved posteriorly at its
anterior end (Fig. 267), no distinct sclerotized tube (ostium bursae) extending distal of the
plates of lamella postvaginallis and lamella antevaginallis, no membranous pouch with
many longitudinal wrinkles on the ventral surface originating at the edge of the lamella
antevaginallis and extending posteriorly over the lamella antevaginallis and genital
opening.
Distribution: Neotropical
Gnathotrichiti


272
amount of codable interspecific variation for Chlosyne. Many taxa also display extensive
intraspecific and/or geographic variation with respect to some elements of the wing
pattern. Long series of specimens from different parts of a taxons range were examined
when coding wing characters, with the exception of certain rare taxa which occupy a
limited range and for which little material was available (see Methods section).
Terminology for wing pattern elements of the Nymphalid ground plan follows
Table 2.1, page 25 in Nijhout (1991). For some wing pattern elements not named in
Nijhout (1991) I have provided names as indicated in Figures 303-309, which reference
elements of the Nymphalid ground plan plus some additional elements not included in Nijhout
(1991) for selected representatives of Chlosyniti and Poladryiti.. The terms "median band" and
"postmedian band" are adopted from Bauer (1961). Also, I substitute "postmedian dots"
for the wing pattern elements homologous to Nijhout's (1991) "border ocelli" because in
the Chlosyniti and Poladryiti most taxa have dots composed of a single color of scales as
opposed to eye spots.
One taxon, Microtia elva, has a wing pattern highly uncharacteristic of any
Melitaeinine. While the genitalia are not unusually divergent, and clearly place it within
the five taxon Microtia clade, its wing pattern is far too divergent to provide evidence
that this taxon belongs in the Melitaeini or of its relationship to other Melitaeinine taxa.
This taxon may be mimetic, and its wing pattern features have apparently evolved at a
much faster rate than other related taxa. In fact, I have no idea how the unusual M. elva
wing pattern features are homologous to elements of the Nymphalid ground plan.
Consequently, 1 have no way of coding most of the wing pattern characters for this taxon.
It is likely that the wing pattern of M. elva is somehow derived from the ground plan
elements (forming a variety of autapomorphic character states), and therefore coding this


107
middle of this assemblage of taxa. In order to preserve the valid taxon Phycioditi,
Melitaeiti must be restricted to a clade of Palearctic taxa including four described genera:
Melitaea, Mellicta, Cinclidia, and Didymaeformia.
The sample of Melitaeiti examined was far to small to test the validity of the four
genera. However, there seems clear evidence from multiple characters that the Melitaeiti
form a monophyletic group, although since not all taxa were examined, there is the
possibility that the some of the synapomorphies given may not be universal. Characters
coded which vary within the Melitaeiti have thus far supported the Higgins (1981)
generic arrangement, although this may well change with the examination of additional
taxa.
Distribution: Higgins (1981) reports the distribution of taxa I place within the Melitaeiti
to include all of Europe, most of Asia excluding the most southeastern areas, northwest
Africa, the northern Arabian peninsula, and Japan.
Gnathotrichiti Subtribe N.
Synapomorphies from binary characters: Orientation of inner valve process anterior
lateral (Figs. 59 & 62), as opposed to posterior lateral in all other Melitaeini (this
character could not be scored for the taxon Gnathotriche mundina which has a vestigial
inner valve process). Posterior ventral inner wall of each valve extended inward in a
flattened, triangular projection with the anterior side approximately perpendicular to the
midline of the genitalia capsule and the posterior side slanted inward (Figs. 58 & 61-64).
Paired saccular projections connected by a broad plate of sclerotized tissue between them
(Figs. 58-59 & 61-64).


60
61. A hollow tube open on both ends on the distal posterior end of the phallus.
l=Absent.
2=Present (Figs.83,88,92,97 & 98).
62. Posterior dorsal end of phallus distal to the posterior ventral side:
l=Entire sclerotized area on posterior dorsal side of phallus in the form of an elongate
triangular tapering area extending distinctly distal to the ventral side of the phallus
opening (Figs. 29-30).
2=Not extending distinctly distal to the posterior ventral end (Figs. 49-57 & 77-79).
3=With an elongate tapering extension with the sides slightly concave (Figs 88,92 & 97-
98).
4=A triangular extension with a squared off tip (Figs. 65 & 71).
5=A broad extension with a convex base (Figs. 104,127,129,131,133 & 213-233).
63. For those taxa with state 5 of the preceding character, the presence of a triangular projection
at the distal posterior edge of the phallus extension:
l=Present (Fig. 104).
2=Absent (Figs. 127,129,131,133 & 213-233).
Independent evidence from other characters suggests state 5 of the preceding character is a
terminal derived state with one independent acquisition in G. exclamationis, therefor taxa lacking
this state are coded 0and G. exclamationis is coded ?.
64. For those taxa with an extension of the dorsal posterior side of the phallus distal to the ventral
posterior side, the sclerotization pattern on the ventral side of this extension:
l=Sides and middle darkest (the sides may be as dark or darker than the middle) with the
sides and middle separated by an area of lightly sclerotized tissue of variable width (best
illustrated in Figs. 1 & 228, difficult to see in some figures of state 1 taxa because the


443
extent of variation within local populations and metapopulations. For some taxa museum
series contained large series of a taxon showing considerable variability, but with a
substantial portion of the specimens containing only vague locality data. For some
species or subspecies, I have cited reports of local variation in the literature which I was
not able to independently verify.
Sexual Dimorphism: Most Chlosyniti have little sexual dimorphism other than average
differences in size. A separate section describing sexual dimorphism is included for taxa
with a marked sexual dimorphism throughout their range. If sexual dimorphism exhibits
geographic variation in part of a taxon's range, it is discussed under that section. If only a
few minor characters vary or tend to vary between sexes, this is noted where those
characters are discussed in the descriptions.
Range: I summarize the distribution of specimens I have examined, and then note
records in the literature (if any) for areas from which I have not seen any specimens.
Distributional data based on specimens which I have examined versus literature sources
are clearly distinguished for all taxa. Table 12 presents a chart of data from museum
specimens which I have examined, with the most detailed coverage provided for taxa
which are relatively uncommon or occur primarily south of the United States border.
Extensive distributional data for butterflies of North America north of Mexico are already
available from the distribution maps in Scott (1986) and Stanford and Opler (1993), and
several additional sources for particular species, although I provide additional records
outside the United States/Canada distributional area for some of these species that extend
their range into Mexico or beyond. For those taxa with a primarily Canadian and/or U. S.
distribution, I make note of records outside of the ranges indicated by Scott (1986) and


Table 12 Continued
742
NMNH=Nat Mus, of Natural Hist.-Smithsonian Inst FSCA=FL St Collection of Arthropods, AM=Allyn Mus., PC=Personal Collection; @=Judged by author as of dubious authenticity
Country
State/Prov-
County (USA)
Locality (non US)
Locality (USA)
Locality Continued (non USA)
Date
Chlosyne erodyle poecile
Chlosyne erodyle poecile
Chtoy>e erodyle poecHe
Chlosyne erodyle poecile
Chlosyne erodyle poecile
Chlosyne erodyle poecile
Chlosyne erodyle poecile
Chlosyne erodyle poecile
Chlosyne erodyle poecile
Chlosyne erodyle poecile
Chlosyne erodyle poecile
Chlosyne erodyle poecile
Chlosyne erodyle poecile
Chlosyne erodyle poecile
Chlosyne erodyle poecile
Chlosyne erodyle poecile
Chlosyne erodyle poecile
Chlosyne erodyle poecile
Chlosyne erodyle poecile I
Chlosyne erodyle poecile I
Chlosyne melanarge (NMNH.FSCA)
Chlosyne melanarge
Chlosyne melanarge
Chlosyne melanarge
Chlosyne melanarge
Chlosyne melanarge
Chlosyne melanarge
Chlosyne melanarge
Chlosyne melanarge
Chlosyne melanarge
Chlosyne melanarge
Chlosyne melanarge
Chlosyne melanarge
Chlosyne melanarge
Chlosyne melanarge
Chlosyne melanarge
Chlosyne melanarge
Chlosyne melanarge
Chlosyne melanarge
Chlosyne melanarge
Chlosyne melanarge
Chlosyne melanarge
Chlosyne melanarge
Chlosyne melanarge
Chlosyne melanarge
Chlosyne melanarge
Chlosyne melanarge
Chlosyne melanarge
Chlosyne eumeda (NMNH,AM,FSCA)
Chlosyne eumeda (M)
Chlosyne eutnede (M)
Chlosyne eumeda (M)
Chlosyne eumeda (F)
Chlosyne eumeda (melanic M)
Chlosyne eumeda
Chlosyne eumeda
Chlosyne eumeda (M)
Chlosyne eumeda
Chlosyne eumeda (M)
Chlosyne eumeda
Chlosyne eumeda (F)
Chlosyne eumeda (melanic M)
Chlosyne eumeda
Chlosyne eumeda
Chlosyne eumeda
Chlosyne eumeda (M)
Chlosyne eumeda (M.F)
Costa Rica
Costa Rica
Costa Rica
Costa Rica
Costa Rica
Panama
Panama
Panama
Panama
Panama
Panama
Panama
Panama
Panama
Panama
Panama
Colombia
Colombia
Colombia
Venezuela
Mexico
Mexico
Mexico
Mexico
Mexico
Mexico
Mexico
Guatemala
Guatemala
Guatemala
Guatemala
Guatemala
Guatemala
El Salvador
Nicaragua
Nicaragua
Nicaragua
Nicaragua
Nicaragua
Nicaragua
Nicaragua
Nicaragua
Costa Rica
Costa Rica
?
Colombia
No Data
Mexico
Mexico
Mexico
Mexico
Mexico
Mexico
Mexico
Mexico
Mexico
Mexico
Mexico
Mexico
Mexico
Mexico
Mexico
Mexico
Mexico
Mexico
Guanacaste
Guanacaste
Puntarenas
Puntarenas
San Jose
Canal Zone
Canal Zone
Canal Zone
Canal Zone
Canal Zone
Canal Zone
Chiriqui
Veraguas
Aragua
Guerrero
Oaxaca
Oaxaca
Oaxaca
Retalhuleu
Managua
Managua
Managua
Managua
Managua
Guanacaste
Guanacaste
Cojima
Colima
Colima
Guerrero
Guerrero
Guerrero
Guerrero
Guerrero
Guerrero
Guerrero
Guerrero
Guerrero
Guerrero
Guerrero
Guerrero
Guerrero
Michoacan
Michoacan
Aphelandra sp (Acanthaceae)
Nr Plata Del Cocos
Buenos Aires {
Villa Colon |
Canal Area. Ft Kobbe, ex larva o
Canal Area. Ft Kobbe, ex pups
Canal Area. Ft Kobbe, ex pupa
Canal Area. Ft Kobbe. ex pupa
C Z Paraso
Paraso
Santa Cruz
Ballena
Henry Pittier Nat Park. Maracay. along mountan stream in ran forest 3500'
Acapulco
Candelaria Loncha [_
Candelaria Loxicha
Candelaria Loxicha
Candelaria Loxicha [
Candelaria Loxicha I
43 Ml W of Tehuantepec
Escuintlai
Retalhuleu
Zac apa
4 Mi N Santiago de Maria
Managua
10 Ml SW Managua
15 km S Managua I
12 MIS Managua
Granada | _
Granada 1
5 Mi SE Granada
6 Ml S. 6 Ml W Canas, 10 19'N 85 09 W
8 km NW Bageles, Comelco Prop., 100m
Com3i3. 2100 fl
La Salada. 1000 ft
La Salada. 1000 ft
AcaftutooBo (sp?)
Acahuizotla (sp?)
Acapulco

Sierra de Guerrerc
23 Sept 1973
20-VI-1976
1-27-71
10-VI1-1971
18-VI-1980
9 Jan 1985
2-1-85
34-85
29-XII-85
19-VI-1978
VII-14-77
VIII-10-1975
1-3 II 1979
5-10-70
8-9-70
8-15-70
Nov
17 Sept 1968
20 Sept 1968
4 Sepl 1969
6 Sepl 1969
22 Sept 1969
22 VIII 63
18-23 Jun 1986
At
15-VII-1971
29 Aug 1972
23 Aug. 1975
22 Oct 1976
21 July 1974
19 Oct 1975
16X1 1967
23 Aug 1975
31 July 1976
23 Aug 1975
10 Jul, 1968
VI-23-73
8-VI1-1968
115-VI-1968
; 10-VII-1968
VI-56
VHI-64
9-17-67
Aug 06
! VI-57
lvit-57
vm-S7
IVI1-59
¡VIII-60
Sepl
¡Aug. 10
j VII 1950
¡VIII 1951
E J Garbetg
J K Batcrunas
H L King
GB Small
G B Smril
GBSmJ
GB Smd
G B Small
G B Small
G B Smart
G B Smart
G B Smart
HLKing
H L King
HLKing
B Neumogen
Edw T Owen
W Schaus
Robbins
HV Weems Jr
Schaus & Barnes
P JLandoll
Schaus & Barnes
HLKing
Collection Brklyn Mus
GF&SHevel
R Anderson
R Anderson
R Anderson
3 Anderson
E L Todd
} Anderson
R Anderson
R A Anderson
T aboga
Of*
i Neumoegen
S Cas sine
toben Wind
Robert Wind
Robert Wind
W M Schaus
Escalante
' Escalante
Escalante
' Esc at ante
Muller
Collection Brklyn Mus
FSCA
FSCA
FSCA
FSCA
NMNH
NMNH
NMNH
NMNH
NMNH
NMNH
NMNH
NMNH
NMNH
FSCA
FSCA
FSCA
NMNH
NMNH
NMNH
NMNH
NMNH
FSCA
FSCA
NMNH
FSCA
NMNH
FSCA
NMNH
FSCA
NMNH
FSCA
NMNH
NMNH
NMNH
NMNH
NMNH
NMNH
NMNH
NMNH
NMNH
NMNH
NMNH
FSCA
NMNH
NMNH
NMNH
NMNH
FSCA
NMNH
NMNH
NMNH
NMNH
Chlosyne eumeda (F)
Mexico Michoacan Cosfcuayana
VI 1952 T Escalante AM 1
Chlosyne eumeda
Mexico Nayarit Vic Compostela
5-IX-1932 NMNH 1
Chlosyne eumeda
Mexico Sinaloa, Nayarit. Coahuila, OR Chihuahua: San Bias
7-22-56 Mai Douglas FSCA 1 1
Chlosyne eumeda
Mexico ? S Jose de Guaymas
April 10 Howard NMNH l
Chlosyne eumeda (M)
Mexico ? Pur B A J
Chlosyne hylaeus (AM.[NMNH.FSCA])
Chlosyne hylaeus
Mexico Guerrero Mexeaia
VIII-52 T Escalan le AM 1
Chlosyne californica (NMNH.FSCA)
Chlosyne californica
JSA Arizona Coconino Little Springs
Chlosyne californica
ISA Arizona Maricopa Phoenix. Desert Vista
Chlosyne californica
JSA Arizona Maricopa Rio Verde Mts. Phoenix
Aug 1893 W Barnes NMNH ii
Chlosyne californica
JSA Arizona Maricopa Seven Springs
8 April 1988 Robert C Godefroi IfSCA 1
Chlosyne californica
JSA Arizona Maricopa Seven Springs
8 April 1990 Robert C Godefroi FSCA 2
Chlosyne californica
USA Arizona Maricopa Seven Springs
9 April 1990 j Robert C Godefroi FSCA [ 1
Chlosyne californica
USA Arizona Mohave
Chlosyne californica
USA Arizona Mohave
Aug 24-31 ¡NMNH 1
Chlosyne californica
USA Arizona Mohave
Chlosyne californica
USA Arizona Yuma Yuma
Chlosyne californica
USA Arizona ?
Chlosyne californica
!SA California Riverside Chino Canyon
Chlosyne californica
USA California Riverside Chino Canyon
Chlosyne californica
'SA California Riverside Palm Springs. Chino Canyon
Chlosyne californica
SA California Riverside Chino Canyon
10 April 1966 J B Heppner FSCA j 2
Chlosyne californica
USA California Riverside Chino Canyon
Chlosyne californica
JSA California Riverside Palm Springs
March 24-30 Barnes NMNH t
Chlosyne californica
USA California Riverside Palm Springs
Chlosyne californica
SA California Riverside Palm Springs
Chlosyne californica
SA California Riverside Palm Springs
Chlosyne californica
USA California Riverside Palm Springs
Chlosyne californica I
USA California Riverside Palm Springs
Chlosyne cafornica ¡
USA California Riverside Palm Springs
Chlosyne californica
USA California Riverside
Chlosyne californica
USA California Riverside
3-8-31 'FSCA | i


402
89
f
FW eqv.
of 118
FW
Numbers refer to
character numbers
from Table 5
b: Basal band of basal symmetry system
c: Distal band of basal symmetry system
d: Proximal band of central symmetry system
e: Discal spot
f: Distal band of central symmetry system
g: Basal parafocal element
h: Postmedian dot (=border ocellus)
i: Distal parafocal element
j: Submarginal band
k: Marginal band
MB: Median Band
BML: Basal Median Line
DML: Distal Median Line
PMB: Postmedian Band
B: Background
(b-k are based on Nijhout 1991)
Terminology not in Nijhout (1991)
Figure 307: The Nymphalid Ground Plan wing pattern elements on the ventral surface of Chlosyne
kendallorum, and other wing pattern terminology.
eqv.
Pringe
PMB


710
ones mixed in on the ventral side; ventrally with base white and black hairs and scales in
the center and throughout the terminal segment, with the white lateral bands extending
down along the sides; inner lateral side with white scales and hairs (some black may be
mixed in) dorsally and at base and black scales and hairs ventrally and at tip (some
specimens may have sparse white scales mixed in at the tip). Vertex with a distinct white
centered patch reaching the posterior edge of the plate, only black scales elsewhere.
Sutures at the lateral edges of the vertex with black scales only. Tuft of white hairs on
the frontoclypeus anterior-lateral of each antennal base usually well developed.
Frontoclypeus with black scales and hairs as background, also with white patches of
scales touching the eyes ventro-laterally of the antennal bases and a white patch centered
on the ventral side of the face. Patch of orange scales absent at the edge of the inner
anterior-lateral margin of the eye. Black and sometimes white scales present in the collar
between the head and pronotum. Abdomen with dorsal-lateral white bands at the
posterior edges of abdominal segments which may be continuous or incomplete and
composed of only scattered light scales, black scales elsewhere. Ventrally the
longitudinal parallel black stripes are present against a yellow-cream background, with
the stripes appearing broken between segments in some specimens. Tibia and tarsi of
prothoracic legs variable, in the nominate janais form individuals may have orange scales
dorsally and white scales and hairs ventrally, or have white scales dorsally, a thin band of
orange scales on the basal-lateral side, and black scales and hairs ventrally. Specimens
examined of the marianna and gloriosa phenotypes all had the former condition, while
intermediate specimens were found with both conditions. Femur of meso- and
metathoracic legs with predominately orange scales dorsally and black at the proximal


515
cream or off white background color distal to this area. Ventral hindwing without heavy
black scaling along the veins but with thin black scaling along parts of veins.
Geographic Variation: C. nycteis exhibits the same range of variation between light and
dark individuals as described above for C. harrisii, and like in C. harrisii, I find no gaps
in the range of variation of any character and consequently recognize no subspecies.
There appears to be a slight increase in the tendency for a greater amount of orange
between the bands of the central symmetry system moving north to south, however the
relative amounts of brown and orange in this area appears to be variable throughout most
of the range (where series of specimens were available). Similar to the situation with C.
harrisii, Manitoba series include individuals which are on the average the lightest, with
all individuals examined at or near the light end of the continuum of variation. However,
individuals near at the light end of the continuum of variation also occur in at least
Wisconsin and Michigan, and close individuals occur throughout the range where long
series were available (most eastern and midwestem states). Series from throughout the
range, excluding Manitoba, include a mix of darker and lighter individuals. Bauer (1975)
applied the name drusius to individuals "with an increased black-brown pattern on the
upperside...The underside has slightly darker coloring and a more complete pattern"
reported to occur in Wyoming, Colorado, Arizona, New Mexico, and Texas. However, I
did not find individuals which I have examined from Texas, Colorado, and New Mexico
to be outside the normal range of variation, although I have seen only darker individuals
with predominately brown between the symmetry system bands from New Mexico and
Colorado. The name reversa has been applied to the light populations from the Riding
Mountains, Manitoba and vicinity (Bauer 1975, Scott 1986).


621
specimens from the pale cream end of the continuum or like some of the intermediate
fulvia phenotypes.
Females do exhibit geographic variation in the amounts of orange and black on
the dorsal wing surfaces. Specimens from Kansas, Utah, and Colorado include females
that are as orange as the alma. The darkest females from Kansas are as dark as the
lightest females from Arizona, but many Arizona females are darker than the range of
variation exhibited by Kansas females and none are as orange as the alma phenotype of
C. leanira leanira. The average amount of dark in females seems to progressively
increase in populations moving from north to south. However, I have seen to few
specimens from Texas, Oklahoma, or Nevada to make generalizations from those areas.
Another character that varies along the same continuum as the amount of black is the
amount of contrast exhibited by the cream colored markings on the dorsal wing surfaces.
The orangest specimens have only the postmedian bands markedly contrasting, with the
remaining cream markings pale and diffused with orange (except for the forewing median
band in the anterior most cells). The darkest specimens have all of the cream dorsal
markings contrasting markedly, with little or no orange diffusion although a variable
amount of orange may surround the periphery of some cream markings against a black
background, especially in the patch at the distal end of the discal cell. Every intermediate
condition is exhibited along the continuum from the orangest to darkest individuals, but
there is also some variation in the degree of orange diffusion within cream markings and
the amount of orange surrounding the cream patch at the distal end of the discal cell even
within specimens with a comparable amount of black. The ventral forewing is slightly
variable in how contrasted light and dark markings are within the washed out orange


40
taxa with the lobe have spines, there would seem a good chance that these characters are
not independent.
21. For those taxa with state 2 of the preceding character, the spine like setae are:
l=Numerous, dozens are present (Figs. 2,3 & 15).
2=Only five, especially large ones are present (Fig. 6).
Independent evidence from other characters suggests state 2 of the preceding character is
a terminal derived state; therefore taxa lacking this state for the preceding character are
coded 0.
22. The shape of the area containing dense setae on the inner posterior side of the valve:
1=A roughly bean-shaped plateau with a flat surface (except for setae sockets).
2=The area is curved and not forming a plateau.
3=The area forms a plateau with a broad ventral side, widening slightly and then
tapering to almost a point on the dorsal side.
4=The area forms a plateau with a narrow, rounded ventral side and a broadly
rounded dorsal side.
All Melitaeini have an opening between where the plates of a valve fold around
and overlap, always visible in ventral view. In most taxa, this opening ranges from not
visible to barely visible along the ventral edge in lateral view. However, the position of
the opening is notably different in Euphydryiti.
23. Position of the above ventral valve opening in lateral view:
l=Completely and clearly visible in lateral view dorsal to the ventral edge of the
valve (Fig. 15).


383
independent data sets should yield primarily groupings which either agree or do not
conflict with each other, and that neither data set should produce well supported
groupings that conflict with the other data set. Comparison of the proportion of groups in
conflict between two data sets, and the proportion in conflict from one data set with both
data sets combined, can also test the relative effectiveness of different methods, such as
the UPGMA algorithm versus parsimony, equal versus successive weighting of
characters, and eventually (if a CALEOS algorithm is developed) analysis with PAUP's
"multistate taxa treated as polymorphisms" option versus a strict consensus of the
CALEOS and distinct state models.
Since selection acts upon the entire phenotype, and phenotypic traits may
represent a compromise between conflicting selective pressures (Bell 1997), arguably no
morphological characters are completely independent in the absolute sense. However, I
consider it likely that genitalic characters and pattern characters constitute somewhat
independent data sets( i. e., the form of a wing pattern component is more likely to affect
the evolution of another wing pattern component to a grater extent than for a sclerotized
structure of the genitalia).
Summarizing the results for equal character weighting and parsimony, ten clades
were present in both data sets, all clades (100%) in one data set were compatible with
those in the other (in agreement, or resolved in one and unresolved in the other), and all
clades in the tree derived from both data sets were in agreement with all of the clades
obtained from separate analyses, plus some additional clades were obtained from
combined analysis. I interpret this as extremely strong evidence supporting the
theoretical assumptions used in this study to reconstruct phylogeny. When I generated


25
Since the Phycioditi and Eurasian Melitaeiti are large groups, with 137 and 67
species, respectively, listed in Higgins (1981), and since the aim of this study was only to
determine if these groups are monophyletic and where their clade appears on the
Melitaeinine phylogeny, I limited my coverage of these groups to a sample. I coded
characters for 18 Phycioditi, including the type species of each genus. I also examined
Higgins (1981) Figures 183-477, which include genitalia drawings of males of most of
the Phycioditi, and while these figures are limited in their detail, they do suggest the
sample of taxa examined provides good representation of the range of genitalic variation
within Phycioditi. I suspect most or all characters and character states not included in the
sample are particular to certain Phycioditi, and consequently not important to an analysis
aiming only to determine if the Phycioditi are monophyletic and the position of their
clade on the Melitaeinine tree. I coded characters for 13 species of Eurasian Melitaeiti,
including the types of each genus plus several additional representatives.
The number of male individuals which I dissected varied among the in group taxa.
For taxa examined in the genera Chlosyne, Thessalia, Charidryas, Anemaca, Texola,
Dymasia, Microtia, Antillea, and Poladryas, I dissected a minimum of three individuals
from different parts of a taxons range, and in nearly all cases more than three individuals
were examined. The sole exception is Texola anomalus, for which I was only able to
examine one specimen. For Phycioditi and Eurasian Melitaeini, two or more individuals
were examined for the type species of each genus, and one to two individuals for the
remaining taxa. For Euphydryiti, three or more individuals were dissected for Nearctic
taxa and the types of each genus, while one to two individuals were dissected for
remaining taxa. I dissected only one individual of Gnathotriche sodialis and


511
Diagnosis: Chlosyne nycteis is sometimes misidentified as C. harrisii, and occasionally
as Chlosyne gorgone or Phyciodes tharos. The separation of C. nycteis and C. harrisii is
presented under C. harrisii above. Chlosyne nycties is most easily separated from C.
gorgone by examination of the ventral surface, where the unique arrow-head shaped
sections of the distal median line and submarginal band will readily separate C. gorgone.
There are many wing characters by which to distinguish P. tharos, but among the most
obvious include the absence of ventral orange forewing and hindwing bands between the
submarginal and marginal bands and the absence of the contrasting silvery white area on
the ventral hindwing.
The male genitalia of C. nycteis are similar to other members of its clade. The
posterior side of the inner valve process is concave in ventral view (Fig. 134) versus
convex in C. harrisii (Figure 136). In contrast to C. gorgone, C. nycteis lacks spines on
the posterior valve process. A difference between C. nycteis and C. kendallorum is less
clear, although no representatives of C. nycteis were found to have a saccus shaped like
C. kendallorum (Figure 137), but only one male specimen of C. kendallorum was
examined and the saccus shape varies considerably within many Chlosyne.
Further Description: Both orange and black scales and hairs on the labial palpi.
Dorsally, palpi with predominately black and scattered orange and white scales and hairs,
outer lateral side with an orange band (also including a variable number of black scales)
dorsal to a white band, ventrally with a mosaic of black hairs mixed in with white ones
and a white base, inner lateral side with black scales and hairs dorsally and a mosaic of
black and orange scales and hairs ventrally. Vertex with black and orange scales.
Sutures at the lateral edges of the vertex with black, orange, and white scales. Tuft of


42
great diversity of valve structure and setae arrangements within the Nymphalini and
Kallamini not found within the Melitaeini.
27. Posterior ventral inner wall of each valve extended inward in a flat, triangular
projection with the anterior side approximately perpendicular to the midline of genitalia
and the posterior side slanted inward.
l=Absent.
2=Present (Figs 58 & 62-63).
28. Proximity of the valvae to each other in ventral and dorsal view:
0=Valvae distinctly separated (Fig. 1 and all other figures of male genitalia
capsules in ventral or dorsal aspect).
l=Valvae touching each other or nearly so (Figs. 34-35).
The only taxon that was somewhat ambiguous in scoring this character is Mellicta
aurelia. In this taxon the valvae are still distinctly separated, but not by as much as other
state 0 taxa.
29. The presence of a patch of short setae on the dorsal half of the posterior lateral sides
of the valvae (distinct from the long setae at the posterior edge of the valvae):
0=Present (Figs. 15,36,39,42,45,64.2,76,82,96,101,114, & 176-192 (minus
178,179,183 7 187).
l=Absent (Figs. 60,64.1,116,118 & 120).
This character can be difficult to detect with a camera lucida setup, and it was
inadvertently omitted from some of the first figures I produced of lateral genitalia
capsules, including Figs. 91,178,179,183 & 187. These taxa actually do have the patch of


115
concave posterior side and convex anterior side. Posterior end of valvae terminating in a
short, curved, pointed tooth or a long tubular extension with a flared flattened end.
Sockets for setae on inner sides of valvae visible in ventral view. A patch of short setae
is present on the dorsal half of the posterior lateral sides of the valvae anterior to the area
with dense hair like posterior setae. Valvae widely separated in ventral view. The
ventral valve opening is located more anterior than posterior and closest to the inner edge
of the valve. Surface of juxta smooth, and juxta lacking projections. Saccus forked with
a pair of invaginated projections with their openings concealed or exposed by the
vinculum, or unforked with a single invaginated process located centrally or off side from
the midline. Tegumen expanded into a well developed uncus. Uncus simple or with
paired claw-like projections or multiple teeth located on the posterior sides. Ventral
posterior side of phallus lightly sclerotized in the middle and more heavily sclerotized on
the sides, with the lightly sclerotized area extending anterior of the supersensory
membrane. Dorsal posterior side of phallus uniformly sclerotized, and not extending
much beyond the ventral edge, although a posteriorly attached laterally flattened structure
connected to the dorsal surface of the everted vesica is present in some representatives.
The proximal plate and distal teeth-bearing plates on the vesica are present, with the latter
in the form of an arched plate with prominent teeth spread over its surface, with some
teeth distinctly larger than others. Vesica otherwise simple, lacking teeth or granulse
patches.
Females: Sclerotized plate on ventral corpus bursae lacking anterior extensions and
inverted teeth. Inverted teeth on corpus bursae nonsclerotized and concolorous with the
corpus bursae or very lightly sclerotized, arranged in two disjunct patches on the lateral


444
Stanford and Opler (1993), and of parts of the distribution maps of these and other
sources for which I have not examined voucher specimens. In some cases, I find need to
explicitly question the authenticity of the labels on some specimens which I have
examined and of some literature records, and provide an explanation for doing so in all
such cases.
Species/Subspecies Delimitation: This section explicity states the evidence upon which I
have based decisions regarding delimitation of species and subspecies taxa. Evidence
considered includes the phylogenetic evidence provided in the preceding chapter,
distributional data of sympatric or allopatric occurrence, discontinuities in morphological
variation, and in a few cases biological information provided by the literature. The
theoretical and applied species and subspecies concepts upon which my decisions are
based are discussed in the introductory chapter.
Key to the Genera, Species, and Subspecies of the Chlosvniti
My philosophy regarding dichotomous keys is that a key should be designed to
overall facilitate accurate identification of taxa as easily as possible. I do not feel it is
important that keys follow the phylogenetic arrangement of taxa. In some cases, the key
incidentally follows the phylogenetic order, while in others it does not because I view an
alternative arrangement as easier for identification purposes. For example, a key that has
Microtia grouped together would have to rely on genitalic characters, when identification
of Microtia species is readily possible without genitalic dissection. Also, when possible I
have preferentially used characters that are easiest to observe on a spread specimen
without genitalic dissection, although in some cases I supplement these characters with
genitalic characters.


340
Variation in the Proportion of Homoplastic Characters with Different Numbers of
Taxa in the Analysis
The number of parsimony informative characters, the number of homoplastic
characters, and the proportion of characters which are homoplastic above various nodes is
presented in Table 9. These values are presented separately for genitalic and pattern
characters in addition to for all characters combined. As above, node numbers refer to
the clades designated in Figure 312. The proportion of homoplastic characters versus
number of taxa in the analysis for genitalic, pattern, and all characters in presented in
Figure 332.
As would be expected, there is a general pattern of an increase in the proportion
of homoplastic characters with an increase in the number of taxa included in the analysis.
For genitalic characters, there were no instances of homoplasy (with the exception of one
character in the Charidryas clade, yielding a value of 33.3% for the proportion of
homoplastic characters because only three parsimony informative genitalic characters
were present in this clade) until the number of taxa in the analysis reached 30, while for
pattern characters instances of homoplasy began to appear with as few as eight taxa.
Between eight and 30 taxa in the analysis, the proportion of homoplastic pattern
characters underwent a steep increase, between 30 and 48 taxa the value fluctuated with
no pattern of increase, and then at the largest node (all Poladryiti and Chlosyniti),
including 54 taxa, the proportion of homoplastic pattern characters reached its highest
value at 65.9%. The proportion of homoplastic genitalic characters, barring the
Charidryas clade with one out of three informative characters homoplastic, never


9
that I recognize as basal taxa are equivalent units, because divergence between clusters
varies in kind and degree. In my view, it is essential that taxonomists first objectively
describe the patterns of variation that they find in nature, and then translate this
information into species/subspecies decisions, as opposed to basing descriptions of the
patterns of variation in nature on an attempt to conform to a particular species/subspecies
concept.
The Theoretical Species Concept Used in this Work
Although I find the goal of a universally applicable clear-cut species concept
untenable, I do conclude that one of the species concepts proposed has the greatest utility
to the field of biological systematics. In my view, a species concept useful to biological
systematics should have two important characteristics: 1) All species units are natural
(=monophyletic) taxa; and 2) Species units have an additional biologically and
evolutionarily significant characteristic that distinguishes them from higher and lower
monophyletic groups, such as genera and populations, respectively. However, with any
species concept one must just accept that there are some cases where lineages will be in a
state of transition in the speciation process, and in these cases species taxa delimitations
will have to be made more arbitrarily.
The species concept that in my view best achieves the above goals is the
evolutionary species concept described by Wiley (1992), where species are defined by:
"A species is a single lineage of ancestral descendent populations of organisms which
maintains its identity from other such lineages and which has its own evolutionary
tendencies and historical fate" (Wiley 1992). I regard a more concise definition as: a
species is a monophyletic lineage biologically incapable of reticulating with a different


364
kinds of polymorphisms with overlapping derived states (such as A=l&2, B=2&3, and
C=3&1. If the first character in the doublet would be coded 1 for A or C but it would be
unclear what should be coded for B), but this scenario did not occur in the
Chlosyniti/Poladryiti data set.
Scenarios 2 and 3 (Figures 338 & 339) illustrate the circumstances where
CALEOS and DPCWH models will produce different results (Scenario 3 illustrates the
same point as Scenario 2 but merely adds one additional complexity). With respect to
scenario 3, consider an ancestor which had state 0 for a particular character which gave
rise to four taxa. One of the descendant taxa has state 0 (A), two have state 1 (C and D)
and one has states 0&1 (B). Under the CALEOS model (A(CBD) would not be shorter
than an alternative topology because any topology would require two steps, for example
(CBD) could involve an acquisition of 1 followed by a loss of 0, or an independent
acquisition of 1 and change from 1 to 0 could occur, yielding a topology ((AB)(CD)).
Under the DPCWH, the latter topology would be 0.5 steps longer. I would argue that
parsimony does not imply that (A(CBD)) is shorter than ((AB)(CD)) if the 0&1
polymorphism is due to genetic differences (of course if 0&1 is actually a distinct state
taxon B would require a change of 1 step wherever it is placed on the tree). If one change
occurred at the ancestor which gave rise to (CBD) at least one additional change must
have occurred subsequently, either the loss of the genetic basis for producing state 0 or its
reacquisition. If no assumption is made that a change from 0 to 1 requires more steps
than a change from 0 to 0&1, ((AB)(CD)) can be accounted for by two steps (as well as
(A(BCD))), a change from state 0 to 1 in the (CD) ancestor and the independent
acquisition of the ability to produce state 1 in taxon B. In effect, the DPCWH model is


241
However, the variation in the saccus can be divided into three distinct states. Most taxa
are assigned state 0 below, while the other two states are limited to two taxa each, which
fall far outside the considerable range of variation found among and within state 0 taxa.
33. Invaginated extensions of the saccus:
0=Two bilaterally symmetrical projections, one on each side (Figures 80,94,
99,108,110,112,122 & 134-155).
#l=Saccus tapering anteriorly and triangular, then forming a narrow extension
forked with two small extensions at the anterior most end (Figures 87 & 89).
#2=Saccus roughly triangular with a bulge at about one-half its length, slightly
forked at the anterior most end (Figure 112).
Juxta:
34. Ventral surface of the juxta:
#0=Smooth with no raised plateau or ridges (Figures 80,87 & 89).
1=A prominent raised plateau composed of a diamond shaped posterior section;
the plateau is constricted at the posterior end of the diamond section, and widens
anteriorly into a broad triangular-shaped section with slightly concave sides. The
lateral sides of the raised plateau are steep and distinct while the anterior and
posterior sides more gradually slope dorsally to the level of the remainder of the
juxta (Figures 108 & 134-155).
2=A distinct posteriorly directed triangular-shaped plateau, including the entire
anterior part of the juxta as the base of the triangle; the sides and tip are bordered
by steep dorsal slopes; the apex of the triangular plateau may or may not extend
as a narrow, pointed posterior projection (Figures 110 & 122).


101
projection. Granulse patches on tegumen conspicuous or obscure at 50X magnification.
Lightly sclerotized area on ventral posterior side of phallus extends anterior to the
supersensory membrane. Vesica lacking sclerotized plates, and with the band of
granulse patches longer than wide.
Females: Inverted teeth on corpus bursae conspicuous at 37.5X magnification with good
lighting.
Hypodryas Higgins
Synapomorphies from binary characters: Distal end of inner valve process flared out and
serrate (Figs. 10-14).
Terminal derived states from multi state characters: Mid ventral juxta ridge terminating
distinctly before the anterior edge of the juxta (Fig. 5). Inverted teeth on corpus bursae
distinctly smaller and more difficult to detect relative to other Euphydryiti (this character
is homoplastic within the Melitaeini).
Further description: Males: Distal edge of the dorsal fork of the inner valve process
narrowly rounded off or sharply pointed, of variable orientation, with or without
prominent teeth on one surface but when present the surface bearing teeth is orientated
posteriorly. Base of dorsal process either broad or narrow. Spine-like setae on inner
valvae walls numerous. Inner posterior wall of valvae with many hair-like setae
smoothly curved over or forming a plateau with a broadly rounded dorsal side and
narrowly rounded ventral side. Dense, minute, granulse structures present on the ventral
midline and other parts of the juxta in all but one taxon. Posterior midline of juxta
extends as a laterally flattened, two-piece posterior extension, slightly more heavily
sclerotized than the remainder of the juxta. Posterior tegumen projections vary in shape


794
Mayr, E. 1963. Animal Species and Evolution. Harvard Univ. Press, Cambridge,
Mass.
Miller, L. D., and F. M. Brown. 1981. A catalogue/checklist of the butterflies of
America north of Mexico. Lepidopterists Society Memoir No. 2.
Miller, L. D., and B. Rotger. 1979. Two new Chlosyne (Nymphalidae) from Mexico.
Bulletin of the Allyn Museum, No. 54: 1-4.
Mishler, B., and M. Donoghue. 1992. Species concepts: a case for pluralism. In
The Units of Evolution by Marc Ereshefsky. MIT Press. Cambridge, Mass. Pp.
121-138.
Neck, R. W. 1980. Aberrant specimen of Chlosyne lacinia from central Texas
resembles topical form. J. of the Lepidopterists Society, 34(4): 363-364.
Nijhout, H. F. 1991. The Development and Evolution of Butterfly Wing Patterns.
Smithsonian Institution Press, Wash., D. C.
O'Hara, R. J. 1994. Evolutionary history and the species problem. Amer. Zool.,
34:12-22.
Opler, P. A. 1999. A new Chlosyne checkerspot from northeast Mexico (Lepidoptera:
Nymphalidae: Nymphalinae). Holarctic Lepidoptera, 6(1): 23-24.
Opler, P. A., and G. O. Krizek. 1984. Butterflies East of the Great Plains. John
Hopkins University Press, Baltimore, MD.
Priestaf, R. C., and J. F. Emmel. 1998. A new subspecies of Thessalia leanira
(Lepidoptera: Nymphalidae) from the coastal sand dunes of central California. In
Systematics of Western North American Butterflies, Thomas C. Emmel, Editor.
Mariposa Press, Gainesville, FL. Pp. 139-158.
Riley, N. D. 1975. A field guide to the butterflies of the West Indies. William Collins
Sons & Co. Ltd. Great Britian.
Robbins, R. K. 1991. Evolution, comparative morphology, and identification of the
Eumaeine butterfly genus Rekoci Kaye (Lycaenidae: Theclinae). Smithsonian
Contributions to Zoology No. 498. Smithsonian Institution Press, Wash., D.C.
Robbins, R. K., and P. M. Henson. 1986. Why Pieris rapae is a better name than
Artogeia rapae (Pieridae). J. of the Lepidopterists Society, 40(2): 79-92.
Scotland, R. W. 1992a. Character coding. In Cladistics: A Practical course in
Systematics by Peter J. Forey et al. Oxford University Press, New York.
790


426
Variation in Mean Retention Index for Genitalic Characters, Pattern Characters, and all
Characters with Different Numbers of Taxa in the Analysis
Mean Genitalic Character Rl i Mean Pattern Character Rl % Mean Rl for all Characters
Figure 330: Variation in mean retention index for genitalic characters, pattern characters, and all
characters with different numbers of taxa in the analysis, based on the Chlosyniti/Poladryiti data matrix
(Table 5) analyzed with heuristic searches with multistate taxa treated as polymorphisms in PAUP 4.0b4a


417
Bootstrap
100
C. nycteis
C. gorgone
C. n. sp. nr. har.
C. harrissii
C. hoffmanni
C. palla
C. gabbii
C. acastus
C. w. damoetas
C. w. whitneyi
C. definita
T. ezra
T. theona
T. perlula
T. chinatiensis
T. I. leanira
T. I. fulvia
T. cyneas
T. cynisca
C. e. pardelina
C. e. endeis
C. melitaeoides
C. marina
C. e. poecile
C. e. erodyle
C. melanarge
C. eumeda
C. californica
C. lacinia
C. ehrenbergi
C. hippodrome
C. narva
C. g. gaudealis
C. g. wellingi
C. janais
C. marianna
C. rosita rosita
C. montana
C. rosita browni
C. riobalensis
C. mazarum
T. eleda
T. coracara
T. anomalus
D. dymas
M. elva
A. pelops
A. proclea
P. minuta
H. fasciatus
H. miriam
A. perezi
A. pantoni
A. tulita
CumPhyGnaMel
Figure 322: Boot strap 50% consensus tree for successively weighted characters from the Chlosyniti/
Poladryiti data matrix, with the DPCWH method for characters with multistate taxa.


248
50. A ventral extension of the outer edges of the lamellae antevaginallis beyond the
plane of the rest of the lamella antevaginallis, forming concave depressions on each side:
0=This area is fairly smooth, lacking such extensions.
l=The above extensions are present (Figure 283).
All Chlosyniti and Poladryiti have a lightly sclerotized area on the lamella postvaginallis
around the opening to the ductus bursae.
51. The continuity of the lightly sclerotized area at the lamella postvaginallis base with
the posterior edge of the lamella postvaginallis:
0=Non contiguous (Figures 277-281,284-288 & 289-292).
l=Narrowly contiguous (Figure 283).
2=Very broadly continuous (Figure 282).
52. Anterior edge of lamella antevaginallis with a strongly posteriorly curved broad
emarginate process overlapping the lamella postvaginallis for most of its length and
enclosing the ductus bursae opening in a pouch. Dorsally the process has posterior
extensions fused with the lamella postvaginallis on each side of the base which form
walls around the genital opening.
0=Lacking a process.
#l=With the process described above (Figures 274-275).
Pattern Characters
Body and Appendages:
The Chlosyniti exhibit a diverse array of interspecific variation with regard to the
patterns and color of scales and hairs on parts of the body and appendages. It is difficult
to know if all of this variation is completely independent, but the characters coded were


155
'23(2),24( 1)
41(2),42(2).
39(1),40(1)'
Figures 58-60. Male genitalia capsule of Gnathotriche exclamcitionis. 58: Ventral. 59: Dorsal


676
patch, and both features become progressively larger moving along the continuum
between quehtcila and lacinia until they are fused together into one large patch.
Form saundersi is most similar to form adjutrix. It consistently differs from
adjutrix by having a broad dorsal hindwing orange patch like form lacinia rather than the
orange median band and black basal area present in adjutrix. Also, consistently the
orange and cream scales on the dorsal hindwing are not differentiated into separate
patches (if any cream scales are present) in saundersi unlike adjutrix. The ventral
background color tends to be a darker more yellowish cream in saundersi relative to
adjutrix. The light patches basal to the dorsal forewing median band are often larger in
saundersi relative to adjutrix. It is common for saundersi to have well developed dorsal
hindwing orange patches distal to the postmedian dots (apparently more common in
South America than Central America) while I have seen these feature only occasionally
in adjutrix relative to the number of specimens examined. In South American specimens
of saundersi the dorsal forewing black areas basal to the median band are usually
extensively diffused with orange (unlike adjutrix) but this condition seems to be
uncommon and less marked in Central American representatives of saundersi. Form
saundersi integrades to form lacinia, form quehtala, and to form paupera. It co-occurs
with either quehtala or lacinia integrades in some areas or both. In these areas, form
paupera is apparently absent. However, it appears form paupera and sautulersi/paupera
intermediates are present wherever form saundersi occurs without lacinia or quehtala
integrades (South America and Trinidad), but regular saundersi is more numerous in
collections. Form lacinia specimens integrading to saundersi do not exhibit further
modification of the dorsal hindwing pattern relative to form lacinia, and the specimens


362
I argue that either the distinct state or the CALEOS model is the best way to
analyze character state polymorphisms, and that knowledge of which is most appropriate
depends on knowing whether the polymorphism is due to genetic differences in
individuals within a taxon or to the presence of a fixed allele or group of alleles that
produces more than one phenotype. Consequently, it is often impossible to know which
model would be best to use for most morphological characters with multistate taxa since
this information is almost always unknown. Therefor, 1 argue that ideally a data matrix
with polymorphic characters would be analyzed with both models, and then that clades
not common to both be discarded unless evidence is acquired to prefer one model over
the other for the relevant character(s). Scenarios 1 and 5 illustrate situations where both
models produce the same result, and in my view yield information content from character
state polymorphisms which is phylogenetically valid with the assumptions of either
model but unobtainable with PAUP 4.0b4as treatment of polymorphisms.
It is interesting to note that with the distinct state model the average minimum
number of extra steps required for a character state not representing a polymorphism is
0.2545 (n=444), while the same average is 0.2500 for a character state representing a
polymorphism (n=44). These averages were calculated with autapomorphic states
excluded, since such states have no possibility of contributing to homoplasy on any tree
topology. Consequently, in this case study on the average distinct states representing
character state polymorphisms contribute to homoplasy about equally to other character
states. However, under the CALEOS model some of the distinct state models character
states representing unique polymorphisms would require a homoplastic step.


646
has an orange tint in C. erodyle poecile, in contrast to males of C. melitcieoides. The
ventral hindwing band of black (fused basal parafocal elements) basally bordering the
postmedian band patches is basally to distally as wide or wider than the width of these
patches in C. erodyle poecile, while it is distinctly narrower in C. melitcieoides. The
ventral forewing and hindwing patches between the position of the submarginal band and
distal parafocal element are always large and conspicuous in C. melitcieoides and often
reduced (sometimes only conspicuous in cell M3, and small even there) in C. erodyle
poecile. A number of body and appendage pattern characters separate C. melitaeoicles
from C. erodyle poecile, with some of the most apparent including the palpi and legs.
The ventral base of the labial palpi is white in C. melitaeoicles but black in C. erodyle. C.
melitaeoides has only orange scales on the tarsi and tibia of the meso- and metathoracic
legs while C. erodyle poecile has a mix of black and orange, often with black dominating.
Further Description: With the exception of the characters noted in the diagnosis of C.
marina, the characters in the detailed description of C. marina also are applicable to C.
melitaeoides. Wing span 36-42mm c?s, 41-43mm $s.
Geographic Variation: C. melitaeoides appears to occupy a restricted range and 1 have
seen no evidence of geographic variation, granted the number of specimens examined is
not large.
Range: I have examined specimens only from Mexico, including the states of San Luis
Potos (Road between Cardonas and Ciudad Del Maiz and Santa Catarina), Nuevo Leon
(Montemorelos-Rayonas), Hidalgo (Jacala), and Tamaulipas (San Fernando). Table 12
includes specimen data from the NMNH, AM, and FSCA. Ferris (1989) credits C.
melitaeoicles to the United States, with no further locality data.


211


Table 12 Continued
749
NMNH=Nat. Mus, of Natural Hist.-Smithsonian Inst.; FSCA=FL St. Collection of Arthropods; AM=Allyn Mus.; PC=Personal Collectio
g=Judged by author as of dubious authenticity
Country
County (USA) Locality (USA)
Locality (non US) Locality Continued (non USA)
Collectior no
Chlosyne lacinia
uq Panama
?
8.27.70
H L King
FSCA
T
Chlosyne lacinia
s.sp i Colombia
Antioquia
Medellin, Sunflower
Gallego M
NMNH
4.1
Chlosyne lacinia
p
Colombia
Atlntico
S A Cotton, F L Gallego
NMNH
1
Chlosyne lacinia
s
Colombia
Caldas
Mun de Palestina, Region de Santaqueda. Host Sunflower
Aug 1970
J Adriano Padilla
NMNH
5
Chlosyne lacinia
s
Colombia
Distrito Especia
Bogota
NMNH
3
Chlosyne lacinia
s
Colombia
Guajira
Nazareth |illegible|, 1000'
25 May 1970
James H Baker
NMNH
t
Chlosyne lacinia
s
Colombia
Valle del Cauca
C8i
28 July 1973
R W Hutchings
!fsca
|
Chlosyne lacinia
s
Colombia
Valle del Cauca
Cali
27 July 1977
R W Hutchings
FSCA
,
Chlosyne lacinia
s
Colombia
Valle del Cauca
Cali
28 July 1977
R W Hutchings
FSCA
1
Chlosyne lacinia
s
Colombia
Valle del Cauca
Cab
30 July 1977
RW Hutchmgs
FSCA
,
Chlosyne lacinia
s
Colombia
Valle del Cauca
Cali. 3000'
30 June 1981
J Bolling Sullivan
FSCA
1
Chlosyne lacinia
s
Colombia
Valle del Cauca
Cali. 3000'
8 Jan 1985
J Bolling Sullivan
FSCA
3
Chlosyne lacinia
s
Colombia
Valle del Cauca
1 Cbi, 3000'
18 Jan 1986
J Bolling Sullivan
FSCA
2
Chlosyne lacinia
s
Colombia
Valle del Cauca
Cali, 3000'
IS Jun 1988
J Bolling Sullivan
FSCA
2
Chlosyne lacinia
s
Colombia
Valle del Cauca
El Retiro
23 May 1981
R W Hutchings
FSCA
1
Chlosyne lacinia
s
Colombia
Valle del Cauca
Hormiguero, near Cali
19 Jan 1986
J B Sullivan
FSCA
,
Chlosyne lacinia
s
Colombia
Valle del Cauca
¡ Loboguerrero
12 Jan 1985
J B Sullivan
FSCA
2
Chlosyne lacinia
s
Colombia
Valle del Cauca
Loboguerrero
130 Jan 1988
J B Sullivan
FSCA
2
Chlosyne lacinia
s
Colombia
Valle del Cauca
Rio Anchicaya
8-III 1964
NMNH
1
Chlosyne lacinia
s
Colombia
[?
Corumba
March
NMNH
,
Chlosyne lacinia
s
Colombia
?
NMNH
3
Chlosyne lacinia
s
Venezuela
Carabobo
Valencia, reared on Helianthus annus
Sep 23, adult Oct.
C H Ballou
NMNH
a
Chlosyne lacinia
s
Venezuela
Falcon
Maganta Id. El Morro
13 Feb. 1985
J F G Clarke 8 N L McIntyre
NMNH
ii
Chlosyne lacinia
s
Venezuela
Falcon
Maganta Id La Sierra. 2000'
25 Feb. 1989
J F G Clarke 4 N L McIntyre
NMNH
,
Chlosyne lacinia
s
Venezuela
Falcon
Maganta Id La Sierra. 2000'
27 Feb. 1989
J F G Clarke 6 N L McIntyre
NMNH
2
Chlosyne lacinia
s
Venezuela
Falcon
Maganta Id La Sierra, 2000'
2 III 1989
J F G Clarke 4 N L McIntyre
NMNH
,
Chlosyne lacinia
s
Venezuela
Falcon
Nueva Esparata Isla Maganta Las Piedras. 11 00'N 63 54W. 600 ft
115 Aug. 1987
J Glassberg
NMNH
,
Chlosyne lacinia
s
Venezuela
Federal District
Caracas
27 Dec 1938
Bro Anthonius
NMNH
2
Chlosyne lacinia
s
Venezuela
Federal District
Caracas, ex Helianthis annus, 9.17'41
@Em. X-1'48
C H Ballou
NMNH
,
Chlosyne lacinia
s
Venezuela
Federal District
Caracas. X 17-41. pupa IX 23
Em. X-l
NMNH
Chlosyne lacinia
s
Venezuela
Federal District
Caracas. Helianthis annus, IX.17'41-42, pupa IX.23
Em. XI
A. T. B
NMNH
,
Chlosyne lacinia
s
Venezuela
Federal District
Caracas, 3200 ft
26 March 1970
Robbins
NMNH
,
Chlosyne lacinia
s
Venezuela
Federal District
Caracas, 4200 ft
29 March 1970
Robbins
NMNH
1
Chlosyne lacinia
s
Venezuela
Federal District
Caracas
E H Pittier
NMNH
,
Chlosyne lacinia
s
Venezuela
Federal District
S of Caracas. 4000 ft
111-31-70
NMNH
,
Chlosyne lacinia
s
Venezuela
Federal District
40 mi W of Caracas. 1800 ft
1 April 1970
Robbins
NMNH
Chlosyne lacinia
s
Venezuela
Federal District
El Valle, Ex Sclerocarbus coffeaecola
27 Jul 1938. adult 19 Aug. 1938; C H Ballou
NMNH
,
Chlosyne lacinia
s
Venezuela
Federal District
El Valle, Ex Sclerocarbus cofeaecola
j 27 Jul 1938, adult 22 Aug. 1938; C H Ballou
NMNH
,
Chlosyne lacinia
p
Venezuela
Federal District
: El Valle, Ex Sclerocarbus coffeaecola
27 Jul 1938, adult 22 Aug 1938; C H Ballou
NMNH
3
Chlosyne lacinia
S,p
Venezuela
Lara
El Cuj
128-IX-80
A J Escalona
FSCA
Chlosyne lacinia
s
Venezuela
Merida
^ D .'Vflhnlrtnn
NMNH
Chlosyne lacinia
s
Venezuela
?
B Neumogen
NMNH
1
Chlosyne lacinia
s
Venezuela
j?
EH Pittier
NMNH
,
Chlosyne lacinia
s
Venezuela
?
NMNH
,
Chlosyne lacinia
s
Trinidad
St. George East
Arima Rd. mile post 17
9/9/79
T C Emmel
FSCA
,
Chlosyne lacinia
s
Trinidad
La Brea
9/11/79
T C Emmel
FSCA
8
Chlosyne lacinia
s
Trinidad
Moruga
E W Rotor
NMNH
,
Chlosyne lacinia
s
Trinidad
Xeres Felo
23 Oct. 1943
W H Wagner Jr.
NMNH
Chlosyne lacinia
s
Peru
Apurimac
Abancay. el. 7200'
123 Feb 1970
James H Baker
NMNH
,
Chlosyne lacinia
S/P
Peru
Leoncio Prado
Two different locality labels
22-VH982
FSCA
1
Chlosyne lacinia
s
Peru
Loreto
Pucallpa
12-VIIH961
A Schmitt
NMNH
7
Chlosyne lacinia
s
Peru
Loreto
Urubamba R 2500'
24 May 1970
James H Baber
NMNH
Chlosyne lacinia
S.p
Ecuador
(Not recorded)
NR
Kieth WiDmott
! Pers Col
2.1
Chlosyne lacinia
s
Bolivia
Santa Cruz
Santa Cruz Savedra Exp. Sta., 430m, Ex: Sunflowers reared (Giraso 1)
! Mil-1977
H Serrate
NMNH
7
Chlosyne lacinia
p
Bolivia
?
Yungas |mote than one state has "Yungas" in the name
¡IV-54
J C Hop finger
NMNH
,
Chlosyne lacinia
s
Brazil
Mato Grosso
Chapada
March
NMNH
Chlosyne lacinia
s
Brazil
Mato Grosso
Chapada
EdwTOwen
NMNH
1
Chlosyne lacinia
s
Brazil
Mato Grosso
Corumba
March
NMNH
2
Chlosyne lacinia
s
Brazil
Mato Grosso
Corumba
EdwTOwen
NMNH
Chlosyne lacinia
s
Brazil
Mato Grosso
B Neumogen
NMNH
Chlosyne lacinia
s
Brazil
Minas Genas
Nova Lima. 900m, 19 59'N 43 52W
121 April 1991
Robbins & Becker
NMNH
,
Chlosyne lacinia
s
Brazil
1 Rio Grande do Sul
Guarani
4 III 1932
C M Biezanko
NMNH
,
Chlosyne lacinia
s
Brazil
Rio Grande do Sul
Pelotas
3 I 1954
C M Biezanko
NMNH
Chlosyne lacinia
s
Paraguay
Paraguari
Sapucay
W T Foster
NMNH
Chlosyne lacinia
s
Argentina
| Chaco
Resistencia, as larva on leaves of Xanthium strumarium
Mar/17/90
NMNH
,
Chlosyne lacinia
s
Argentina
Formosa
RA
Dec 1917
W M Schaus
NMNH
,
Chlosyne lacinia
s
Argentina
La Rioja
E Giacometti
NMNH
j
Chlosyne lacinia
s
Argentina
Misiones
San Ignacio
i V-27-1961
Krauss
NMNH
,
Chlosyne lacinia
s
Argentina
Tulman
Tucuman Chaar, ex Larva lee aves of Tithonia tubaeformis (Compositae)
17 V 1983
NMNH
Chlosyne lacinia
s
Argentina
1?
May 1920
NMNH
~
Chlosyne ehrenbergi (NMNH.FSCA
Chlosyne ehrenbergi
Mexico
Guerrero
Taxco
12-30-53
FSCA
,
Chlosyne ehrenbergi
Mexico
Federal District?
Atzcapco |=Atzcapotzalco?|
31 Aug 1922
Chlosyne ehrenbergi
Mexico
Federal District
Trotrvaca Rumas
30-IX-1967
H L King
¡FSCA
Chlosyne ehrenbergi
Mexico
Jalisco
Guadalajara
7 Aug 1957
G W Rawson
FSCA
,
Chlosyne ehrenbergi
Mexico
Jalisco
Guadalajara
¡27 Aug 1957
,
Chlosyne ehrenbergi
Mexico
Jalisco
Guadal a) ara
W Schaus
NMNH
Chlosyne ehrenbergi
Mexico
Jalisco
Guadalajara
NMNH
Chlosyne ehrenbergi
Mexico
Jalisco
Ocotlan
Chlosyne ehrenbergi
Mexico
Jalisco
Ocotlan
8 Oct '52
S S Nicolay
NMNH
Chlosyne ehrenbergi
Mexico
Jalisco
Ocotlan
¡ 15 April '53
,
Chlosyne ehrenbergi
Mexico
Jalisco
Tlaqueplaque
IIX-1965
N L H Krauss
¡NMNH
Chlosyne ehrenbergi
Mexico
Michoacan
Tar ecu ato
23X71
ILeeuw
FSCA
Chlosyne ehrenbergi
Mexico
Michoacan
Jruapan
May 1900
Chlosyne ehrenbergi
Mexico
Morelos
Ex. Cuernevaca
XI-1944
N L Krauss
NMNH
Chlosyne ehrenbergi
Mexico
Morelos
Cuernavaca
IX-1965
NLH Krauss
NMNH
Chlosyne ehrenbergi
Mexico
Morelos
Cuernevaca
June 06
W Schaus
NMNH
Chlosyne ehrenbergi
Mexico
Morelos
Cuernevaca
Chlosyne ehrenbergi
Mexico
Morelos
Popocatepetl Pak, 8-1000 ft
Jul '06
|W Schaus
NMNH
,
Chlosyne ehrenbergi
Mexico
Morelos
Popocatepetl Park
W Schaus
Chlosyne ehrenbergi
Mexico
Morelos
Morelos St.
J C Hopftnger
I NMNH
Chlosyne ehrenbergi
Mexico
Morelos
Tejafpa
¡9-IX-1967
Chlosyne ehrenbergi
Mexico
Oaxaca
Ejutla. on MX 175
8/16/80
EC Olson
FSCA
Chlosyne ehrenbergi
Mexico
Oaxaca
usl outside of Ejutla on Mex 175
16 Aug 1980
EC Olson
Chlosyne ehrenbergi
Mexico
Oaxaca
Oaxaca
¡15 Aug.'57
,
Chlosyne ehrenbergi
Chlosyne ehrenbergi
Chlosyne ehrenbergi
lxico
Oaxaca
Oaxaca
Oaxaca
Otaca
25 Aug.-57
¡FSCA
n
Mexico
Oaxaca
_jCB Richards
W Schaus
I NMNH
i
Chlosyne ehrenbergi
Mexico
Oaxaca
20-25 Mi S of Oaxaca
8/25/74
Mike Bell
FSCA
,


266
Colorado have extensive black scales on the femur whereas C. damaetas xvhitneyi from
California have very few.
75. The pattern of scales on the femur of the meso- and metathoracic legs.
0=Orange scales dorsally, white scales ventrally (except at distal end where scales
all orange).
l=Predominantly orange scales dorsally except black scales at proximal end,
predominantly black scales ventrally except orange at distal end.
*2=Covered with black scales.
3=Covered with black scales except orange in distal most end.
#4=Covered with orange scales only (some Atlanteci specimens may also have
sparse off-white scales).
*5=As in state 1, except a row of grey scales borders the white scales.
@*A=0&1: Geographically variable (see above).
@*B=0&4: Individuals may be either state 0 or state 4.
@*C=0&5: Some individuals are regular state 0 while others have a unique state
(based on a sample of n=4).
Some specimens of Chlosyne endeis have a few black scales at the base of the femur but
are otherwise like state 0.
76. Scales on the tibia of the meso- and metathoracic legs.
0=Orange scales dorsally, white scales ventrally
l=Covered with orange scales.
2=Covered with black/gray scales.
*3=A mix of black and orange scales with black scales most numerous.


109
This may prove to be an additional terminal derived state for the subtribe after females of
the other taxa are examined. Other characteristics of the G. exclamationis female are as
follows.
Sclerotized plate on ventral corpus bursae lacking anterior extensions and inverted
teeth, narrow and well over five times as long as wide, anterior edge not upcurved.
Inverted teeth on corpus bursae minute, nonsclerotized and concolorous with the corpus
bursae. Ductus bursae vestigial. Ostium bursae absent. Lamella antevaginallis poorly
developed relative to lamella postvaginallis. Lamellae lacking ridges or extensions.
Gnathotriche Felder & Felder
Since I am synonymizing the genus Gnathotrusia Higgins, the description Gnathotriche
is identical to that of Gnathotrichiti.
Remarks: Species in the Gnathotrichiti were placed in Phyciodes by Hall (1928-1930)
and later moved to Gnathotriche by Forbes (1946) (Higgins 1981), where they have
remained since that time although Higgins (1981) split Gnathotriche into two genera, the
other being Gnathotrusia. Higgins (1981) further considered Gnathotriche not to belong
to the Phycioditi, but rather to his paraphyletic concept of Melitaeiti. Higgins (1981)
claimed Gnathotriche and Gnathotrusia showed astonishing similarity in some features
of the male genitalia to Didymaeformia didyma and its allies, suggesting a real
relationship between these taxa, citing the presence of sub-unci (=tegumen projections)
and the shape of the valvae and harpe (=inner valve process) as examples. The posterior
tegumen projections of G. mundina (no tegumen projections are present in G. sodialis
and G. exclamationis) and D. didyma are located in different positions and of vastly
different structure, such that assigning them equivalent states in the data matrix could not


718
meso- and metapleurons posterior to where the femur tucks in as well as small patches of
white just ventral to the wings, and white hairs are usually concentrated around the coxae.
All subspecies have the dorsal forewing black with white markings. The
postmedian dots usually occupy cells CuA2-R5 and R3 on the dorsal surface in
subspecies browni, and usually occupy these same cells on the ventral surface in other
subspecies but are frequently absent in one or more cells on the dorsal surface,
particularly cells CuA2 and R3. Cells CuAl-M2 apparently always have at least small
postmedian dots on the dorsal surface in all individuals. The dorsal forewing median
band usually occupies cells CuA2-R3 in subspecies browni, and may occupy these cells
or be absent in some cells in the other subspecies. Dorsal forewing median band patches
are most likely to be absent in cell M3 followed by cell CuA2, but are apparently always
present (although sometimes small) in other cells. At least small diffuse patches of the
median band are present in cells CuA2-R3 on the ventral surface of all individuals
examined for this feature, and if patches are small and reduced ventrally this occurs in
cells CuA2 and M3 but not other cells. The forewing discal cell almost always has a
white patch basal to the position of the discal spot (only one individual, of subspecies
riobalensis, was found that lacks this patch even on the ventral surface), which is not
divided in contrast to C.jcinais. This spot is almost always present dorsally and ventrally
in subspecies rosita and browni and in both individuals of subspecies mazaran
examined, but is most commonly absent dorsally but present ventrally in subspecies
riobalensis and montana. A second white spot in the anterior part of the discal cell basal
to the position of the distal element of the basal symmetry system is almost always
present in subspecies rosita and browni dorsally and ventrally (also present in both


317
2=Prominent patches in CuA2-Ml (very constant).
3=Variable, patches always occupying CuA2-CuAl and often M3-M2; if a patch
is present in Ml (some individuals of C. rosita browni) it is usually (but not
always in C. r. browni) very small.
Chlosyne rosita browni from the northern part of its range (Texas) tend to have larger
median band patches and more often patches in cell Ml than specimens of C. r. montana
from Mexico and C. r. rosita southward.. The limited material studied for C. riobalensis
(n=5) and C. mazaran (n=2) all were clearly state 3; no individuals had the red median
band extending to cell Ml.
Note for those taxa with character state 2 of the preceding character (coded ?),
the cells occupied by the light orange median band is a highly intraspecifically variable
feature, ranging from cells 1A+2A-Sc+Rl to limited to 1A+2A-CuA2.
125. For those taxa with state 1 of the second preceding character, the most prominent
(largest) patches of the ventral hindwing median band:
l=CuA2-M3 patches (M3 patch rarely absent) distinctly larger than any patches
occurring anterior to these.
2=The largest patches are in cells CuA2 and M2 (in some specimens some other
patches may be similar in size, but never larger).
126. Given the presence of the above ventral hindwing median band (state 1 of the third
preceding character), the presence of this band on the dorsal wing surface:
l=On the ventral surface only.
2=On both the dorsal and ventral wing surfaces.


490
Microtia anmalas (Godman & Salvin)
Phyciodes anmalas Godman and Salvin, 1897. Trans. Ent. Soc. Lond. 1897: 243.
Type Locality: Colima, Mexico.
-Melitaea anmala Rober, 1914. Seitz 5:433. Type Locality: Colima, Mexico.
Diagnosis: M. anmalas may be separated from M. coracara by the presence of the
orange median band noted above. The male genitalia (not illustrated) are like M.
coracara except the specimen examined did not have the openings to the saccular
invaginations exposed, but based on only one specimen I am unsure if this is a consistent
difference.
Further Description: With the exception of the above characters, the description of M.
anomalus is identical to that of M. coracara, although note from Table 5 that some
character states could not be scored due to the condition of the sole representative
examined.
Range: The sole specimen which I have examined is a T. Escalante specimen labeled:
"Coahuayana Mich [Michoacan, Mexico] VIII-50". Higgins (1960) reports the type
locality as Colima, Mexico, and mentions having examined another specimen in the
American Museum but mentioned no associated data.
Species Delimitation: I have some doubts as to whether specific status is truly warranted
for both M. anomalus and M. coracara, given that the only sure difference between them
is a dorsal hindwing median band, which is an intraspecifically variable character in some
Chlosyne. There is no evidence to suggest the two taxa are sympatric, and substantial


629
cyneas found in Arizona also occur far from the known range offulvia in Mexico and
consequently are most unlikely in my view to constitute evidence of hybridization
between cyneas and fulvia.
Chlosyne leanira cynisca (Godman and Salvin)
Phyciodes cynisca Godman and Salvin, 1882. Biol. Cent. Amer. 1:191. Holotype:
British Museum (Higgins 1960).
Diagnosis: C. leanira cynisca is the easiest of the leanira subspecies to recognize. On
the ventral forewing, orange is confined to the discal cell its peripheral areas, and distal to
the postmedian dots. In the other subspecies the ventral forewing is usually
predominately orange, except for some pale individuals of subspecies fulvia which still
have prominent orange distal to the postmedian band. Subspecies cynisca has no orange
on the dorsal wing surfaces. Some individuals of subspecies cyneas have the dorsal wing
surfaces nearly devoid of orange, but usually have at least diffuse orange scaling distal to
the postmedian band. West California coast specimens of subspecies leanira with no
orange on the dorsal wing surfaces occur far from the range of cynisca and could easily
be separated by the ventral forewing and shape of the dorsal forewing median band. In
contrast to the subspecies cyneas, the only leanira subspecies that approaches the range
of subspecies cynisca, C. leanira cynisca has the dorsal forewing median band patch in
cell CuA2 distinctly smaller than the patch in cell CuAl (it may be about the same size or
slightly smaller in cyneas). The dorsal hindwing median band is broad in subspecies


420
. j
S s >
i 8 m
I
1 i
i I
d d
i I
1 *
I I
111 tu]
I i | 1
*411
ill
-i |
i 1
I §
t 4
M i
1 i
d d d d d u
d d d d d
a: 12,25,30,39,52,55,57
61,62,73,75,76,77,85,93,
108,129,135,139
b: 33,34,53,55,69,71,73,
85,90,103,107,108,110,
116,119
d d u u
i 1114 i i i 1 i 1 f
*Polytomies have been removed because MacClade
does not calculate unambiguous changes below polytomies.
Refer to Figure 316 for the strict consensus tree.
Figure 325: Characters that change unambiguously on branch lengths on the most parsimonious tree
derived from the Chlsyniti/Poladryiti data matrix with multistate taxa assigned distinct states.


380
represent taxa. Suppose that for a particular character, the ancestral state is
unambiguously state 0, and taxa 1 and 4 have state 0 while taxa 2 and 3 have state 1.
This character is homoplastic (it has two states but requires two steps, and therefore one
extra step). However, the minimum number of extra steps required by either state 0 or
state 1 is zero. If taxon 4 reacquired state 0, than state 1 is nonhomoplastic and state 0
requires an extra step, whereas if taxa 2 and 3 independently acquired state 1, state 0 is
nonhomoplastic but state 1 requires an extra step. Both alternatives are equally
parsimonious, requiring two steps. Likewise, any time the ancestral node is ambiguous,
it will be impossible to know the minimum number of steps required for character states
which could comprise the ancestral node (this problem not occur for all nonhomoplastic
characters and characters which are homoplastic but for which none of the possible
character states for the relevant ancestral node are homoplastic). This problem could be
addressed in several ways. One is, as I have proposed, to weight states based on the
minimum amount of extra steps they require (character states are not devalued unless
there is unambiguous evidence that they require extra steps to account for their
distribution). This would result in one erroneous weighting in the above example with
taxa 1-4, since both states would be weighted 1 when in fact one of them should be
weighted 0.5. Another possibility is to weight each state as if it is responsible for the
maximum possible number of extra steps required by its character. Likewise, this would
result in one erroneous weighting in the above example, since both state 0 and state 1
would be weighted 0.5 when one of them is nonhomoplastic. Finally, one could base
state weighting on ACCTRAN or DELTRAN models (see Swofford and Madison 1987),
either of which would result in either zero or two state weighting errors in the above


706
=Chlosyne hyperia marianna Rober, 1914. Seitz: 452. Type Locality: Mexico,
Guerrero.
-Chlosyne hyperia irrubescens Hall, 1924. Entomologist. 50:163. Type Locality: S.
Mexico, Cautla.
=Chlosyne gloriosa Bauer, 1959. J. Lep. Soc. 13:165. Type Locality: Tepic, Nayarit,
Mexico, August 16 1954. Holotype: Yale Peabody Museum (Bauer 1959).
Diagnosis: Chlosyne janais is commonly confused with other taxa in collections,
including C. rosita, C. hippodrome, and C. lacinia. The lacinia phenotype of C. lacinia
closely resembles the nominate form of C. janais on the dorsal surface and the quehtala
phenotype of C. lacinia can be confused with the marianna and gloriosa phenotypes of
C. janais. Any form of C. lacinia may be easily distinguished on the ventral surface by
the presence of an orange patch in cell CuA2 extending as far distal as the postmedian
dot. The gloriosa (and marianna) phenotypes of C. janais may be separated from C.
hippodrome by the characters given in the diagnosis of C. hippodrome.
Another source of error regarding the determination of Chlosyne janais that I have
observed in collections lies in distinguishing it from C. rosita. Chlosyne rosita rosita and
Chlosyne rosita hrowni resemble the nominate form of C. janais while C. rosita
riobalensis resembles the gloriosa phenotype. Any subspecies of C. rosita may be easily
and reliably distinguished from C. janais by the absence of ventral forewing and
hindwing patches distal to the postmedian dots; this area is solid black in all subspecies
of C. rosita. C. janais has the characteristic Chlosyne ventral abdominal pattern of two
parallel dark stripes (sometimes appearing broken between segments) against a light


141
Table 2: Status of individual characters used in the heuristic search of the Melitaeini data
matrix with the representative out group method prior to successive weighting
Character
States
Steps
Cl
Rl
RC
Character
States Steps Cl
Rl
RC
1
3
2
1.00
1.00
1.00
61
2
1
1.00
1.00
1.00
2
3
2
1.00
1.00
1.00
62
5
4
1.00
1.00
1.00
3
4
6
0.83
0.90
0.75
63
3
2
1.00
1.00
1.00
4
8
7
1.00
1.00
1.00
64
3
2
1.00
1.00
1.00
5
3
2
1.00
1.00
1.00
65
2
1
1.00
1.00
1.00
6
4
3
1.00
1.00
1.00
66
2
1
1.00
1.00
1.00
7
3
3
0.67
0.90
0.60
67
3
2
1.00
1.00
1.00
8
3
2
1.00
1.00
1.00
68
2
1
1.00
1.00
1.00
9
5
4
1.00
1.00
1.00
69
2
1
1.00
0.00
0.00
10
4
3
1.00
1.00
1.00
70
2
1
1.00
1.00
1.00
11
4
3
1.00
1.00
1.00
71
3
3
0.67
0.94
0.63
12
3
2
1.00
1.00
1.00
72
3
2
1.00
1.00
1.00
13
3
2
1.00
1.00
1.00
73
3
2
1.00
1.00
1.00
14
2
3
0.33
0.91
0.30
74
7
6
1.00
1.00
1.00
15
3
2
1.00
1.00
1.00
75
2
3
0.33
0.75
0.25
16
6
5
1.00
1.00
1.00
76
3
3
0.67
0.86
0.57
17
4
3
1.00
1.00
1.00
77
2
1
1.00
1.00
1.00
18
5
5
0.80
0.89
0.71
78
2
1
1.00
1.00
1.00
19
2
1
1.00
1.00
1.00
79
2
1
1.00
1.00
1.00
20
2
1
1.00
1.00
1.00
80
3
3
0.67
0.96
0.64
21
3
2
1.00
1.00
1.00
81
3
2
1.00
1.00
1.00
22
4
3
1.00
1.00
1.00
82
3
2
1.00
1.00
1.00
23
3
2
1.00
1.00
1.00
83
3
3
0.67
0.89
0.59
24
3
3
0.67
0.86
0.57
84
3
3
0.67
0.96
0.64
25
3
2
1.00
1.00
1.00
85
5
4
1.00
1.00
1.00
26
2
1
1.00
1.00
1.00
86
3 2
1.00
1.00
1.00
27
2
1
1.00
1.00
1.00
87
2
3
0.33
0.50
0.17
28
2
1
1.00
1.00
1.00
88
3
7
0.29
0.75
0.21
29
2
2
0.50
0.80
0.40
89
7
7
0.86
0.96
0.83
30
2
1
1.00
1.00
1.00
90
2
1
1.00
1.00
1.00
31
4
3
1.00
1.00
1.00
91
1
1.00
1.00
1.00
32
2
1
1.00
1.00
1.00
92
2
1
1.00
1.00
1.00
32.1
2
1
1.00
1.00
1.00
93
2
2
0.50
0.93
0.46
33
9
8
1.00
1.00
1.00
94
2
1
1.00
1.00
1.00
34
3
2
1.00
1.00
1.00
95
2
1
1.00
1.00
1.00
35
2
1
1.00
1.00
1.00
96
2
1
1.00
1.00
1.00
36
2
2
0.50
0.90
0.45
97
2
1
1.00
1.00
1.00
37
3
2
1.00
1.00
1.00
98
3
2
1.00
1.00
1.00
38
3
3
0.67
0.89
0.59
99
2
1
1.00
1.00
1.00
39
2
3
0.33
0.78
0.26
100
3
2
1.00
1.00
1.00
40
8
8
0.88
0.90
0.79
101
2
1
1.00
1.00
1.00
41
3
2
1.00
1.00
1.00
102
2
1
1.00
1.00
1.00
42
3
2
1.00
1.00
1.00
103
2
1
1.00
1.00
1.00
43
2
1
1.00
1.00
1.00
104
2
1
1.00
1.00
1.00
44
3
5
0.40
0.75
0.30
105
2
2
0.50
0.93
0.47
45
2
4
0.25
0.89
0.22
106
2
1
1.00
1.00
1.00
46
2
1
1.00
1.00
1.00
107
3
2
1.00
1.00
1.00
47
2
1
1.00
1.00
1.00
108
2
1
1.00
1.00
1.00
48
3
2
1.00
1.00
1.00
109
2
1
1.00
1.00
1.00
49
6
6
0.83
0.95
0.79
110
2
1
1.00
1.00
1.00
50
8
7
1.00
1.00
1.00
111
2
1
1.00
1.00
1.00
51
3
3
0.67
0.90
0.60
112
2
1
1.00
0.00
0.00
52
5
4
1.00
1.00
1.00
113
2
1
1.00
0.00
0.00
52.1
3
2
1.00
1.00
1.00
114
2
1
1.00
0.00
0.00
53
2
1
1.00
1.00
1.00
115
2
1
1.00
1.00
1.00
54
3
2
1.00
1.00
1.00
116
2
1
1.00
0.00
0.00
55
2
1
1.00
1.00
1.00
117
2
1
1.00
0.00
0.00
56
2
2
0.50
0.80
0.40
118
2
1
1.00
0.00
0.00
57
6
6
0.83
0.97
0.80
119
2
1
1.00
1.00
1.00
58
2
1
1.00
1.00
1.00
120
2
1
1.00
0.00
0.00
59
3
3
0.67
0.83
0.56
121
2
1
1.00
1.00
1.00
60
2
1
1.00
1.00
1.00


215
biological systematics. I report results of experiments testing the relationships between
tree scores, character scores, and boot strap scores with different sized monophyletic
groups included in the analysis. In addition, different models for analyzing characters
with multistate taxa were investigated. Finally, I investigated the assumptions inherent to
models based on equally and successively weighting of characters, and propose an
alternative model based on successively weighting character states.
Materials and Methods
Material Examined
The in group taxa examined included to my knowledge all distinct taxa within the
Chlosyniti. I define distinct taxa as operational taxonomic units (OTUs) separated from
other taxonomic units by a distinct gap in the range of variation for at least one
morphological or biological character. Hence, I did not concern myself with how taxa
were classified as species or subspecies when selecting taxa to include in the analysis; my
philosophy is to include every distinct taxon in the analysis and make decisions on what
should be classified as a species or subspecies after a phylogenetic hypothesis has been
obtained. The vast majority of Chlosyniti taxa classified as subspecies by various
authors do not appear in the data matrix because I found no character with a gap in its
range of variation to indicate such taxa were distinct. Obviously, taxa which
represent points along a continuum of variation are not useful for cladistic analysis, since
if no character discontinuities exist among taxa, a phylogenetic hypothesis for their
relationships cannot be derived.


556
Note: Chlosyne definita, C. endeis, C. marina, C. melitaeoides, C. melanarge, C.
erodyle, and C. eumeda all appear to have no consistent differences in male and female
genitalia between them, although it is possible there may be differences in average
capsule size and length of the posterior valve process. The saccus shape is
intraspecifically variable in all of these taxa. This group of taxa does not form a natural
group, and the most parsimonious tree indicates the genitalic type they share evolved no
later than the ancestor which gave rise to the clade of Chlosyne including C. definita as
the basal taxon (Figure 311 node 19), and consequently represents a symplesiomorphy
for these taxa. A tree where all of these taxa form a monophyletic group (based on
equally weighted characters with PAUP's model for polymorphisms) would require a
minimum of nineteen extra steps over the most parsimonious tree. This is a good
example contrary to the popular misconception that similarity in genitalia among insect
taxa implies close relationship. Basic evolutionary theory implies that similarities only
provide evidence of relationships if they are synapomorphies instead of
symplesiomorphies, and genitalic similarities certainly warrant no exception. The most
parsimonious phylogenetic hypothesis indicates divergence from the ancestral form of the
genitalic capsule occurred in the lineage which gave rise to species formerly classified as
Thessalia (Figure 311 node 22) and the ancestor of the lineage including C. lacinia
(Figure 311 node 34).
Chlosyne definita (Aaron)
Figures 141,182,219 & 240.


435
CALEOS Model
A
0
B
0&1
A
0
3 C
&1
+1
2 Steps -0
2 Steps .+1
+ 1
B
O&l
_+0
2 Steps
A
0
B
a
C
I
a>l
2 Steps
0>a
0
B
Distinct State Model
ABC
p
-0>a
0
C
0>1
A
0
C
1
B
a
2 Steps
-0>1
~l>a
2 Steps
0>!
DPCWH Model
A
0
0
A
0
0
B
0
1
C
1
1
D
I
1
A
0
0
1.0 Steps
"a:0>l
:0>l
''b:l>0
1.5 Steps
a:0>l
0
H
b:0>l
I
B D
0 1
1 1
a: 1 >0
01.5 Steps
Scenario 3
:0>1
. _b:0>l
PAUP &
Discrete
DPCWH
CALOS
State
1
2
1 2
1 2a 2b
OG1
0
0
0 0
0 0 0
OG2
0
0
0 0
0 0 0
A
1
0
1 0
1 0 0
B
1
0,1
1 a
1 0 1
C
1
1
1 1
1 1 1
D
1
1
1 1
1 1 1
Figure 339: Behavior of different models for polymorphic characters on three tree topologies for
Scenario 3, with the polymorphic character mapped on the trees. A-C: CALEOS Model. D-F: Distinct
State Model. G-I: DPCWH Model. J: Data matrix for Scenario 3. The polymorphic character
requires two steps on trees A-C in PAUP 4.0b4a.


148
Figs. 26-33. Phallus Characters of Euphydryiti. 26: H. maturna (ventral). 27: Same (lateral w. vesica
everted). 28: E. desfontaini (ventral with vesica). 29: Same (dorsal). 30: E. edilha (lateral). 31: Same
(lateral). 32: Same (dorsal). 33: Same (ventral).


22
With respect to the cumulative out group method, I have polarized characters
within the Melitaeini based on two assumptions. The first is that the sister group to
Melitaeini includes and is exclusively composed of some combination of taxa currently
classified within the Nymphalini and Kallimini. The second assumption is that the
sample of generic types is representative of the character states that occur both within the
true Melitaeinine out group and the in group (Melitaeini). I have made no assumption
that particular taxa or groupings of taxa within the Nymphalini and Kallimini are most
closely related to Melitaeini.
In addition to the use of the cumulative out group method, I developed a second
method for polarizing characters within the Melitaeini (in a separate analysis) which I
term the Representative Out Group Method. Use of the representative out group
method requires that characters are first scored for the cumulative out group method.
From the resulting data matrix, I then selected representative out group taxa, such that the
group of representative taxa collectively includes all character states occurring within the
in group. These taxa were then entered into the data matrix individually rather than
collectively, and specific states were assigned for each taxon in place of the c
designation (invalid for this analysis because c can refer to more than one character
state). I scored some additional characters that are invariant within Melitaeini but for
which there is variation in states among the representative out group taxa. The reason for
doing this was so that all the representative out group taxa would not, evidence
permitting, come out as a unresolved basal polytomy where the number of equally
parsimonious out group permutations would be multiplied by the number of equally


303
0=Present.
l=Absent.
In the out group taxon Atlantea pantoni this character is scored exclusively based on male
specimens, as all females examined were too dark in this area to score.
Border ocelli'VHindwing Postmedian Dots: A number of taxa in the Chlosyniti have
hindwing postmedian dots inside patches of the hindwing postmedian band between the
parafocal element and Nijhouts (1991) unnamed element g (or in the equivalent position
for taxa where a postmedian band is absent). The presence of postmedian dots in the
forewing appears to be independent of the presence of postmedian dots in the hindwing,
as many taxa differ as to the presence/absence and/or the form of the postmedian dots
between the forewing and hindwing. Some out-group in-group taxa have black dots, but
within a small group of Chlosyne the black dots vary from all black to black with a cream
colored center to entirely cream colored. A number of other Chlosyne have sharp white
hindwing postmedian dots. Many taxa are highly variable intraspecifically as to how
prominent or obscure the hindwing postmedian dots are and how many hindwing cells
contain dots.
The hindwing postmedian dots occur on both the upper and under wing surfaces,
but in all taxa examined (except for the out group taxa Higginsius miriam and H.
fasciatus, where the dots are more prominent dorsally, and Antillea species, where the
postmedian dots are only present ventrally) the dots tend to be more prominent on the
under surface, while otherwise similar in characteristics. In some specimens the
hindwing postmedian dots may only be discemable on the underside, and this is
frequently the case in Chlosyne rosita. While there is variation among taxa as to how




161
49(2) 2.1(1)
52(3) / tv
40X: Figs. 80-84 84
50X: Fig. 85
>(2)
85 52.1(2)
Figs. 80-82. Male genitalia capsule of Atlantea pantoni. 80: Ventral. 81: Dorsal. 82: Lateral.
Figs. 83-84. Phallus of A tlatea pantoni. 83: Ventral/lateral with vesica everted. 84: Lateral with vesica
everted. Fig. 85: Male genitalia capsule of Higginsius fasciatus (lateral).


633
with predominately orange scales on the inner side and at tip (the tip may also contain
variable amounts of black) and predominately dark hairs (with some white hairs) over
white scales on the outer side, with the dark hairs not extending to the tip and the base
with only white scales and hairs; inner lateral side with predominately orange scales and
hairs ventrally, base white with a broken to continuous white band transversing segment
two and black scales and hairs dorsally. Orange scales may be present or absent at the
inner anterior-lateral margin of eyes. Meso and meta-pleurons with black scales and
hairs (some white may also be present) in the groove where the femur tucks in. Antennal
shaft checkered black and white with orange scales absent. Ventral basal discal cell
patch formed from the basal band of the basal symmetry system present dorsally (or not
detectable in dark specimens) and present or absent ventrally. Ventral hindwing
postmedian band varies along a continuum between light colored patches with a variable
amount of black basally and distally (and thin black scaling along the veins) to the distal
and basal black of the parafocal elements fused to form a dark band with light patches
inside.
Other characteristics are like those of C. definita with the exception of the
differences noted in the diagnosis section for C. definita.
Species Delimitation: Phylogenetic evidence indicates C. endeis has no sister species,
and there are discontinuities in pattern characters between C. endeis and all other
Chlosyne.
Subspecies Delimitation: Phylogenetic evidence indicates C. endeis endeis and C. endeis
pardelina are sister taxa, and these taxa differ only by two consistent pattern differences
and appear to be allopatric. It is entirely plausible that integrades do occur between


Congruency Between Tree Topologies Derived From Genitalic and Pattern
Characters 382
4 PHYLOGENETIC REVISION AND MORPHOLOGICAL CHARACTERIZATION OF
THE CHLOSYNITI (LEPIDOPTERA: NYMPHALIDAE: NYMPHALINAE:
MELITAEINI) 439
Introduction 439
Materials and Methods 440
Key to the Genera, Species, and Subspecies of the Chlosyniti 444
Antillea Higgins 459
Antillea proclea (Doubleday and Hewitson) 464
Antillea pelops (Drury) 466
The Microtia Bates and Chlosyne Butler Clade 468
Microtia Bates 469
Microtia eleda (Hewitson) 472
Microtia elva Bates 478
Microtia dymas (Edwards) 483
Microtia coracara (Dyar) 486
Microtia anornalus (Godman & Salvin) 490
Chlosyne Butler 491
Chlosyne harrisii (Scudder) 495
Chlosyne kendallorum Opler 507
Chlosyne nycteis (Doubleday) 510
Chlosyne gorgone (Hubner) 516
Chlosyne hoffmanni 521
Chlosyne palla (Boisduval) 529
Chlosyne whitneyi (Skinner) 536
Chlosyne gabbii (Behr) 543
Chlosyne acastus (W. H. Edwards) 549
Chlosyne definita (Aaron) 556
Chlosyne ezra (Hewitson) 564
Chlosyne chinatiensis (Tinkham) 570
Chlosyneperlula (Menetries) 578
Chlosyne theona (Menetries) 586
Chlosyne leanira (Felder & Felder) 598
Chlosyne endeis (Godman & Salvin) 632
Chlosyne marina (Geyer) 636
Chlosyne melitaeoides (Felder & Felder) 644
Chlosyne erodyle (Bates) 647
Chlosyne melanarge (Bates) 653
Chlosyne eumeda (Godman & Salvin) 656
Chlosyne hylaeus (Godman & Salvin) 659
Chlosyne californica (Wright) 661
Chlosyne lacinia (Geyer) 667
Chlosyne ehrenbergi (Geyer) 685
vii


433


375
characters which can be mapped on the strict consensus tree with the minimum number
of steps and 67 characters which require one or more additional steps to account for their
distribution on any of the most parsimonious trees, and (2) that any alternative tree
topology would increase the number of extra steps required, I argue the assumption that
all characters are equally informative towards reconstructing a phylogeny is not
supported by the actual data in this case study.
Each change from one character state to another can be thought of as one step, but
with successive weighting certain steps are weighted more than others, specifically the
steps associated with homoplastic characters. Each character state within a character
contributes to the number of steps for that character, and every character state different
from the ancestral node of the clade represented by the entire data set will require at least
one step on any tree (its acquisition), while the state present at the ancestral node will not
require any steps unless it is homoplastic (reacquired subsequent to its change or loss). If
the assumption inherent to the successive weighting model, that characters with higher
rescaled consistency indices are proportionately more informative, is in fact true, I argue
that the steps required by derived character states requiring more than one step on the
strict consensus tree, or by ancestral character states that are reacquired on the strict
consensus tree, should be weighted proportionately less than states for which this is not
the case.
Table 11 lists all character states present in the Chlosyniti/Poladryiti data matrix,
and the minimum and actual number of steps those states require on the strict consensus
tree (Figure 316). Those states which occur in only one taxon (indicated by an asterisk
(*)) obviously require one step on any conceivable tree topology and are irrelevant to this


58
57. Sclerotization pattern of the ventral side of the phallus posterior to the supersensory
membrane:
l=Elaborate spear-shaped design (Figs. 1 & 213-233).
2=Posteriorly lightly sclerotized all the way across, anterior to this area the darkly
sclerotized area begins along the sides, widening until extending all the way
across forming a concave border (Figs. 26,30,49-51,53 & 104).
3=A narrow band of dark sclerotization on each side, and lightly sclerotized in the
middle (Figs. 65,68,73,79,88,92 & 98-99).
4=Lightly sclerotized area ovoid, with darkly sclerotized area reduced to an
extremely narrow band on each side at one point, and expanding anterior and
posterior to this point (Figs. 127,129,131,133).
5=Similar to state 2 except for a narrow, posteriorly tapering triangular section of
more heavily sclerotized tissue originating at the supersensory membrane, and
the border of between the lightly and darkly sclerotized areas is convex (Fig. 28).
6=Like state 3 except narrow inward anterior slanting extensions of sclerotized
tissue from the sides (just anterior to the plate attached to the vesica) (Fig. 97).
Several taxa, Euphydryas phaeton, E. editha, and the E. anida complex have the phallus
twisted 90 degrees such that the lateral side is equivalent to the ventral side for other taxa.
This information is coded as a separate character, and assigning a new derived state for
the ventral sclerotization pattern would weight the same feature twice. The lateral side of
these taxa is typical of state 2, and they are coded as such. The same applies to character
60 below.


233
Taxa lacking a hollow posterior process on the valvae are coded 0. While
independent evidence from other characters suggests a hollow process evolved
independently in the taxon H. fasciatus, since H. fasciatus has a unique form for the
above character this independent acquisition is not weighted twice. Other than for H.
fasciatus, independent evidence suggests state 0 of the preceding character represents a
terminal derived state.
16. For those taxa with state 2 of the preceding character, the overlapping of the
posterior valve processes:
l=Overlapping (Figures 206-207).
2=Not overlapping.
All taxa lacking state 2 for the preceding character have nonoverlapping processes, with
the exception of C. leanira where the processes overlap at the tip (unlike state 1 taxa
above). Since the processes of non state 2 taxa for the preceding character project in
different directions, the above variation cannot between these states. Consequently, taxa
with a posterior process but lacking state 2 for the preceding character are coded ?,
while taxa lacking a posterior process are coded 0.
17. Given state 1 of the preceding character, the number of points at which the processes
overlap in posterior view:
l=One (Figure 206).
2=Two (Figure 207).
Since there is considerable independent evidence from unrelated characters suggesting
state 1 of the preceding character is a terminal derived state, taxa lacking this state are
coded 0.


225
Nymphalid Ground plan or other wing pattern elements in the Melitaeini could be readily
established for wing pattern variation between taxa, with the exception of the taxon
Microtia elva. M. elva has a wing pattern very uncharacteristic of Melitaeini, and the
pattern is so divergent as to leave no clues of its homology with any wing pattern
elements found in other taxa. Consequently, this taxon was coded ? for almost all wing
pattern characters (but not body pattern characters). Wing pattern characters are
illustrated by the specimens figured in Plates A through V in Appendix A.
Within the genus Chlosyne there is a great deal of uniformity in many features of
the male genitalia, although greater genitalic variation occurs among the other Chlosyniti.
I found few examples of intraspecific variation other than for the exact shape of the
saccus (see Figures 255-260 for examples). Only a very few of the potential species
separation or codable characters pertaining to the shape of the saccus initially considered
within Chlosyne actually held up with additional dissections. Some of the characters
which vary interspecifically within Chlosyne relate to features of the inner and posterior
valve projections, which were examined and have been illustrated in multiple aspects to
permit accurate coding. I did not find any evidence of geographic variation in the
genitalia for any Chlosyniti taxon.
While there is some interspecific variation in the female genitalia within
Chlosyne, and some minor intraspecific variation, I did not find any character of the
female genitalia which could be coded into discrete states within Chlosyne and only a few
characters that could be coded into discrete states within Chlosyniti. Initially, I had
considered coding the extent of the sclerotized plate around the sclerotized inverted teeth


786


366
of the Cl, RI, and RC is primarily for comparing different trees with similar numbers of
taxa, the number of taxa in a particular data set will still be important in how tree
statistics derived from that data set compare with those from other data sets even when
the data sets represent monophyletic groups, since adding additional taxa to the data set
could incidentally increase or decrease the proportion of more divergent taxa in the data
set. While I do feel, despite their limitations, that reporting tree statistics such as the Cl,
RI, and RC has some comparative value for showing the amount of homoplasy in a
character set (these values do decrease with increased homoplasy in a data set) I think it
is more important to consider the support of individual clades (based on the number and
quality of characters supporting and contradicting them, and if possible congruence
between independent data sets) rather than trying to use tree statistics to judge the relative
quality of overall trees. Furthermore, as I will argue in final section of this chapter, a tree
with lower tree statistics is not necessarily more poorly supported relative to a tree with
higher tree statistics, even for comparing two trees with the same taxa but derived from
different character sets.
While Figures 329-332 clearly indicate on the average that pattern characters are
much more homoplastic within the Chlosyniti than genitalic characters (inferring that
they are on the average less stable over evolutionary time), and that the contribution of
pattern characters to homoplasy increases at a much faster rate as more taxa are added to
the analysis relative to genitalic characters, the assumption that a genitalic character
necessarily provides stronger evidence of relationships than a pattern character is not
warranted. Examination of the data in Tables 7 and 8 indicates that many individual


418
Chlosyne
Antillea
Microtia 1 Poladryiti
| g = i | S> 3 If | i a a
? 3 Sr¡ _f i o E 35 a 8 T J!
3
3 = s 3:=
O = e
111 £ - 1 1 a t I § Jj £ 1 8.J -i =g 1II & £ i J 1 § 2 £
4 t4 u i4 £ .-i, sbsbS^'oJS 3 -=T 5 ai ai E S ai ai : : : ; "S ¡LS > -o
CjOOOOdOO(jhHI-hl-HHI-OOS225S<<3:i<< * -3 S S 3 i II ¡ 1 11 5 -i 1 S-! -S I S a
, Jl-n 4*2y';:!53;j_==:i
tu *o £ ra soaj: s mj: <> 73 v o. a. s <2 o. o. ~ =
£ 1-3
u 8 -2 :l s f H .
s
n
si
96
91
96
61
60
77
75
75
74
56
58
59
51
87
83
80
59
57
64
85
89
86
81
84
85
74
63
75
82
75
69
83
62
83
68
83
80
77
87
85
Boot Strap Values (*=Not in 50% consensus)
Top: Multistate taxa=Polymorphism in PAUP
Middle: Multistate taxa=Discrete state
Bottom: DPCWH Method
All: Successively weighted characters with
C. hylaeus included.
64
54
51
51
97
99
10*
65
60
59
77
76
77
100
100
100
70
68
66
100
100
100
75
79
90
76
81
91
63
61
57
67
65
64
Figure 323: Proposed hypothesis for the phylogeny of Chlosyniti and Poladryiti, and the revised
generic arrangement of Chlosyniti.


Table 12 Continued 737


400
Table 11: Data set for a first iteration of successive state weighting, with character RC values included for
1(0)
im
2(0)
2(1)
2(2)
2(3)
3(0)
3(1)
4(0)
4(1)
5(1)
5(2)
6(0)
6(1)
6(2)
WL
7(D
7(2)
8(0)
?!1I
9(0)
9(1)
9(2)
9(3)
9(4)
10(0)
10(1)
10(A)
12(1)
12(2)
13(1)
13(2)
13(3)
14(0)
14(1)
14(2)
15(0)
15(1)
15(2)
15(3)
15(4)
15(5)
15(6)
15(7)
16(0)
16(1)
16(2)
17(0)
17(1)
17(2)
18(0)
18(1)
18(2)
19(0)
19(1)
19(2)
19(3)
20(0)
20(11
20(2)
20(3)
21(0)
21(1)
21(2)
22(01
22(1)
23(0)
23(1)
23(2)
23(3)
23(4)
24(0)
24(1)
25(0)
25(1)
26(0)
26(1)
27(0)
27(1)
28(0)
28(1)
29(0)
29(1)
30(0)
30(1,
31(0)
31(1)
32(01
32(1)
33(0)
33(1)
33(2)
34(0)
34(1)
34(2)
Stale | Cha
Weight Weight
0333
0.5
0.6
0.333 0.6
0.6
1
1
0 656
0 656
0656
0 656
0.329
0329
0 329
Cha
Stale Eo
34(3)
34(4)
34(5)
35(0)
35(1)
36(0)
3§<1)
37(0)
37Q)
37(2)
38(0)
38(1)
39(0)
39(1)
39(2)
40(0)
40(1)
11(0)
41(1)
42(0)
i2(1)
43(0)
43(1)
1^(0)
44(1)
45(0)
45(1)
45(2)
46(0)
46(1)
47(0)
47(1)
48(0)
48(1)
48(2)
49(0)
49(1)
50(0)
50(1)
51(0)
51(1)
51(2)
52(0)
A2(i)
53(0)
1*3(1)
53(A)
55(0)
55(1)
55(2)
55(3)
55(5)
55(6)
55(A)
56(0)
56(1)
56(2)
56(3)
56(4)
56(5)
56(6)
56(7)
56(8)
57(0)
57(1)
57(2)
57(3)
57(5)
57(6)
57(7,
57(8)
57(9)
57(A)
58(0,
58(1)
58(2)
58(3)
58(4)
58(5)
58(6)
58(7)
58(8)
59(0)
59(1)
60(0)
60(1)
60(2)
60(A)
61(0)
61(1)
61(2)
61(3)
61(A)
61(B)
a
Stale
Weight
1
1
1
1
1
1
1
05
1
1
1
1
1
1
t
1
1
1
1
1
1
1
1
1
1
_J
1
1
1
1
1
1
1
1
1
0.5
0 333
J
1
1
05
1
1
1
1
J
0.5
1
05
1
1
1
1
1
1
1
1
1
1
1
1
05
0 333
0.5_
1
0 333
1
1
1
05
1
Cha
Weight
1
1
04
04
1
1
1
1
1
1
1
1
0 333
0333
1
1
1
1
0333
0 333
0333
0 333
028
028
028
0 662
0662
0 662
0662
0 662
0662
0 662
0.51
0.51
051
051
051
051
0.51
051
051
0.577
0.577
0 577
0577
0.577
0.577
0577
0577
0.577
0577
0519
0 519
0519
0519
0519
0519
0519
0519
0 519
0475
0475
0 351
0 351
0.351
0 351
08
0.8
08
0 8
08
08
0 611
62(1)
62(2)
62(3)
62(A)
63(0)
63(1)
61(0)
64(1)
64(A)
65(0)
65(1)
65(A)
66(0)
66(1)
66£)
66(3)
66(4)
66(5)
66(6)
66(A)
66(B)
6Z0)
67(1)
67(A)
67(B)
68(0)
68(1)
68(2)
68(3)
68(A)
69(0)
69(1)
69^)
69(3)
69(4)
70(0)
70(1)
7K0)
ZK1)
7JA)
72(0)
72(1)
72(2)
72(3)
72(4)
72(5)
72(6)
72(7)
72(A)
73(0)
73(1)
73(2)
73(3)
_73(1)
73(5)
73(6)
73(7)
73(A)
74(0)
ZKD
74(A)
ZAP)
75(1)
75(2)
75(3)
75(1)
75(5)
75(A)
75(B)
75(C)
76(0)
76(1)
76(2)
76(3)
76(4)
77(0)
77(11
77J2)
77(3)
77(1)
78(0)
78(1)
78(2)
78(3)
78(4)
78(A)
79(0)
79(1)
79(2)
79(3)
7?(4)
79(5)
79(6)
79(7)
79W
Stale
Ext Step* Weight
1
1
05
05
J
1
1
025
1
_1
025
J
1
0.5
1
0.5
1
1
1
J
J
1
0.5
0 333
1
1
0333
1
1
1
0.5
0.25
1
1
1
1
Cha
Weight
1
0.5
0.5
1
1
1
1
1
1
1
1
1
1
1
1
1
0.5
1
1
1
1
0.5
0.333
1
0.5
0,5
1
1
1
1
1
1
1
0.5
0.5
05
0.611
0.611
0.611
0 611
025
025
0.348
0.348
0.348
0.278
0278
0278
0.671
0.671
0.671
0.671
0.671
0.671
0671
0.671
0.671
0.435
0.435
0.435
0.435
0.503
0.503
0.503
0 503
0~503
0.375
0.375
0.375
0.375
0.375
1
1
0.282
0.282
0282
U_
'X!_
0 75
075
0.75
0.75
0.75
0.75
0.75
075
0.75
0.345
0 345
0345
0.589
0589
0589
0589
0 589
0589
0.589
0589
0589
0429
0.429
0.429
0429
0429
0429
0429
0 429
0429
0429
0438
0438
0438
0438
0.438
0438
0475
0475
0.475
0 475
0475
0475
0475
0475
0 475
79(9)
79JA) _
79(B)
. 79(D)
80(0)
180(1)
, 80(A)
82(0)
; 82(1)
82(2)
' 83(0)
83(1)
83(2)
; 83(3)
83(4)
83(5)
J 84(0)
.84(11 _
3 84(2)
84(3)
I 84(4)
85(0)
I 85(2) _
, 85(3)
85(4)
85(5)
85(6)
3 85(8)
85(9)
85(A)
85(B)
' 86(0)
,86(1)
. 87(0)
; 87(1)
! 87(A)
88(0)
; 88(1)
88(2)
88(A)
' 89(0)
! 89(1)
89(2)
; 89(A)
90(0)
! 90(1)
90(2)
90(3)
90(A)
90(B)
' 90(C)
; 91(0)
.91(1)
i 91(2)
: 91(3)
91(4)
.91(B)
.91(C)
i 92(0)
I 92(1)
. 92(A)
93(0)
! 93(1)
93(2)
' 93(3)
I 93(4)
93(A)
93(B)
1 94(0)
. 94(1)
. 94(A1 _
95(0)
: 95(1)
96(0)
1796(1)
1 97(0)
. 97(1)
. 97(2)
98(0)
; 98(1)
98(A)
99(0)
I 99(1)
99(2)
' 100(0)
' 100(1)
100(2)
*101(0)
101(1)
101(2)
J 101(3)
1101(4)
!101(5)
101(6)
101(A|
Cha
Cha
Min No
Stale Cha
Cha
Mm No Stale
Cha
Weight
Stale
Extii Steps
Weight Weight
Stale
Ext a Steps Weight
Weight
0475
101 (B
0
1 0762
122(1)
1 0.5
06
0475
* 101(C
0
1 0 762
122(2
0 1
06
0475
102(0
0
1 0741
122(3
0 1
06
0 475
0
1 0741
122(4
0 1
06
0 325
102(3
0
1 0 741
123(0
0 1
0 583
0 325
102(4
0
1 0741
123(1
1 05
0583
0325
* 102 (A
0
1 0.741
123(2
0 1
0583
0.5
103(0
0
1 0.829
124(0
0 1
i
0.5
103(1
0
1 0829
124(1
0 1
1
05
103(2
0
1 0 829
124(2
0 1
1

103(3
1
05 0829
124(3
0 1 1
1
103(4
0
1 0829
125(0
0 1
0 556
1
103(A
0
1 0829
125(1
0 1
0 556
1
104(1
0
1 1
125(2
1 r 0.5
0 556
1
104(2
0
1 i-
126(0
0 1
1
1
105(0
0
1 1
126(1
0 1
1
1
105(1
0
1 1
126(2
0 1
1
1
105(A
0
1 1
126(A
0 1
1
1
106(0
0
1 1
127(1
0 1
1
1
106(1
0
1 1
127(2
T
1
1
106(2
0
1 1
127(3
0 1
1
0 606
106(3
0
1 1
128(0
0 1
0878
0606
106(4
0
1 1
128(1
1 0.5
0878
0 606
106(5
0
1 1
128(2
0 1
0878
0 606
106(6
0
1 1
128(3
0 1
0878
0 606
106(A
0
1 1
128(4
0 1
0878
0 606
107(0
1
05 0.327
128(5
0 1
0878
0 606
107(1
2
0 333 0 327
128(6
0878
0606
107(3
0
1 0327
128(7
0878
0 606
107(A
0
1 0327
128(8
i-
0878
0 606
* 107(B
0
1 0327
128(A
I 1
0878
0.259
108(0
1
05 0226
* 128(B
0 1
0878
0259
108(1
3
0 25 0 226
129(0
0 1
0675
028
108(2
2
0 333 0 226
129(1
0 1
0.675
0.28
109(0
0 1 1
129(2
1 [ 0.5
0.675
028
109(1
0
1 1
129(3
0 1
0675
0467
110(0
1
05 0.612
130(0
0 1
0 796
0467
110(1
0
1 0612
130(1
0 1
0.796
0 467
110(2
1
05 0612
130(2
0 1
0.796
0467
110(3
0
1 0612
130(3
1 0.5
0 796
0.125
" ''>
0
1 0612
130(4
0 1
0 796
0.125
110(5
0
1 0 612
130(A
0 1
0 796
0.125
110(6
0
1 0612
130(B
0 1
0 796
0.125
110(7
0
1 0612
131(0
0 1
1
0 646
110(A
0
1 0612
131(1
0 1
1
0 646
110(B
05 0612
132(0
0 1
1
0 646
110(C
0
1 0612
132(1
0 1
1
0 646
Hid
0
1 1
133(0
0 1
063
0 646
111(2
0
1 1
1 11
0 1
0.63
0 646
111(3)
0
1 1
133(2
1 05
063
0 646
112(0
0
1 0675
134(0
0 1
1
0613
112(1
1
05 0675
134(1
0 ] 1
1
0613
112(2
0
1 0.675
134(2
0 1
1
0613
* 112(A
0
1 0675
, U, J
0 j 1
1
0613
113(1
0
1 1
134 (A
o ; 1
1
0613
113(2
0
1 1
135(0
0 1
0787
0613
113(3)
0
1 1
135(1
1 05
0 787
0613
114(0
0
1 i 1
135(2
0 1
0787
j :
114(1
0
1 1
135(3)
0 1
0 787
1
115(0
0
1 0703
135(4
0 1
0 787
i
115(1
1
05 0703
135(A
0 1
0 787
115(2)
0
1 0703
135(B
0 1
0787
1
115(A
0
1 0703
136(0)
0 1
0.7
'
116(0)
0
1 0449
136(2)
1 0.5
0.7
"r'
2
0 333 0 449
136(3)
0 1
0.7
1
116(2)
0
1 0449
136(A
0 1
0.7
1
* 116(3)
0
1 0449
* 136(B
0 1
0.7
i
* 116(4;
0
1 0.449
137(0]
1 05
0 444
0267
117(0)
1
0.5 0 727
137(1)
0 1
0 444
0 267
ii7(i:
0
1 0 727
138(0)
0 1
1
0267
117(21
0
1 0727
'38(1
0 1
'
1
.117(3;
0
1 0.727
138(3)
0 1
'
1
117(A
0
1 0727
138(5]
0 1
1
0 375
118(0]
0
1 1
139(0)
0 1
0 758
0375
118(1
0
1 1
139(1)
0 1
0758
1
118(2)
0
1 1
139(2)
0 1
0 758

118(3)
0
1 1
139(3)
0 1 0.758
1
* 118(4)
0
1 1
139(4)
0 1
0758
1
* 118(5)
0
1 1
139(A
0 1
0 758
1
118(6)
0
1 1
140(0)
0 1 J
1
1
ii9(7:
0
1 1
. '40(i;
0 1
i
0615
118(8]
0
1 1
141(0)
0 1
0375
0615
119(11
0
1 1
141(1]
1 0.5
0 375
0615
119(2)
0
1 1
142(0)
0 1
0438
1
120(0)
0
1 1
142(1)
1 05
0438
1
120(1)
0
1 1
143(0)
0 1
0.571
1
120(2)
0
1 1
143(1 ]
0 1
0571
0 762
120(3]
0
1 1
143(2)
1 05
0571
0 762
* 120(4]
0
1 1
.144(0]
0 1
1
0 762
120(5]
0
1 1
144(1)
0 1
1
0762
* 120(6)
0
1 1
0 762
120(7)
0
1 1
0762
121(0]
0
1 1
0 762
121(1)
0
1 1
0762
122(0]
0
1 06
comparison


545
the basal edge of the basal median line and the distal edge of the distal median line. In
the most extreme C. gabbii specimens some of the cream scaling is also present in the
basal area and between the median lines, while in others these areas are pearly white as
with the patches distal to the postmedian band. I have seen one specimen (from near
Hemet, California) which is a typical C. gabbii except the cream scaling is so reduced
that a dissecting microscope is needed to see it well, but the presence of the cream scaling
along the median lines was prominent to the naked eye in all other specimens examined.
C. acastus often has the ventral background color uniform throughout, although in a
portion of specimens the patches distal to the distal parafocal elements may be a
somewhat lighter or paler shade of white.
I am uncertain if C. gabbii co-occurs with C. acastus, but Stanford and Opler's
(1993) county distributional maps show dots for both species (one map is labeled
"Chlosyne acastus complex") in five southern California counties. C. acastus specimens
from southern California (the "neumoegeni" phenotype) can be distinguished by having
the amount of black on the dorsal wing surfaces greatly reduced, and all of the dorsal
wing markings orange with slight if any contrast between the median and postmedian
bands (the contrast is always marked in C. gabbii).
Chlosyne gabbii is frequently confused with co-occurring C. palla. Within the
California range of C. gabbii, C. palla has the distinctly cream colored ventral hindwing
background color throughout versus bright pearly white in gabbii distal to the postmedian
band. Throughout its range, C. palla lacks as marked of a contrast in the ventral
hindwing background color distal to the postmedian band relative to the remainder of the
hindwing. The same characteristic will separate C. gabbii from C. hoffmanni, in addition


78
0=Unfolded.
l=Folded over on the inner side such that the distal edge of the lamella
antevaginallis is a three-dimensional structure (Fig. 264).
104. Sides of lamella antevaginallis near the base with a sclerotized structure with a
convex outer side folded around the sides of the lamella antevaginallis and attached to the
abdominal sterna ventrally and the lamella postvaginallis dorsally:
0=This structure absent.
l=This structure present (Figs. 262-266).
This and the preceding character were examined by prying the lamellae apart with a
forceps as well as examining them from ventral aspect.
105. The presence of a lightly sclerotized area on the lamella postvaginallis posterior to
the ductus bursae opening:
l=Absent (Fig. 261).
2=Present (Figs. 262-264,270,273-274,277-291 (several illustrations depict taxa
with this state where it can not be seen due to the angle from which the illustration
was produced).
Note that taxa found to have state 1 are also taxa with a sclerotized ductus bursae. These
taxa have the distal posterior side of the ductus bursae lightly sclerotized. It may be
possible that the sclerotized end of the ductus bursae in these taxa was formed from an
invagination of the lamella antevaginallis and lamella postvaginallis sclerites, and that the
presence of a sclerotized ductus bursae and the above character may not truly be
independent.


778
PLATE K
30


3 31V3d
ILL


kD
|7f>
i1-'
UNIVERSITY OF FLORIDA
3 1262 08555 1827


28
analysis with successively weighted characters. Strict consensus trees were calculated
from heuristic searches, and 50% consensus trees were calculated from the boot strap
analyses. Tree scores (consistency index, retention index, and rescaled consistency
index) were calculated in PAUP based on parsimony informative characters only.
All heuristic searches were conducted with characters unordered (in my view
ordering characters almost always introduces additional unwarranted assumptions into
the analysis). Also, all heuristic searches were run to completion and the strict consensus
trees were calculated from all of the equally parsimonious trees. For rooting options, for
the cumulative out group method the make in group monophyletic option was selected.
For the representative out group method, the option root tree at internal node with basal
polytomy was selected.
Boot strap analyses were run with the fast stepwise addition algorithm and based
on 10,000 replications. The option retain groups with frequency >50% was used for
each boot strap analysis. For analyses based on successively weighted characters, the
option sample characters with equal probability but apply weights was selected.
Results
Characters and Character States for a Phylogenetic Analysis of the Melitaeini
(Lepidoptera: Nymphalidae)
Almost all of the below character states occurring within the Melitaeini are illustrated in
the camera lucida genitalia drawings presented in Figures 1-292, except for a few states
particular to Phycioditi and European Melitaeiti which may be valuable to future
investigations of relationships within these clades, but are irrelevant to establishing if
these groups are monophyletic or how they are related to other clades. While at least one


212
g
111
119
121
2A
17.
45
73
100
16
25
109'
123
Chars that change
Unambiguously on branch
80
84
48
89
le If
45
106
18
59
71
101-
103
T4-
12
- 14
-- 15
- 16
-78
I-60
- 62
70
- 71
-93
- 105
1,2,3,13,24
25,33,39,41.55,79
80,81.83,84,86,110
22
Char 32 1 = 122
Char 52.1 = 123
Note: Trichotomies on the strict consensus tree were removed
because MacClade 3.07 does not calculate unambiguous state
changes below polytomies.
Figure 302: Characters that change unambiguously on each branch of the Melitaeini phylogeny.


355
(1993) considered leanira, fulvia, and cyneas to be three different species, and did not
address C. cynisca because it occurs outside the geographic area they considered. Bauer
(1975) ranked all four as separate species. Smith and Brock (1988) considered cynisca
and cyneas to be subspecies of the same species taxon, and considered leanira and fulvia
to be separate species. A variety of subspecies names have been proposed for C.
leanira and C. fulvia. However, I find none of them to represent distinct taxa and that all
forms of C. leanira and C. fulvia represent points along a continuum of geographic
variation that exhibits a small gap only between but not within leanira and fulvia.
The results of the phylogenetic analyses provided convincing evidence to indicate
C. leanira, C. fulvia, C. cyneas, and C. cynisca form a monophyletic group; however,
evidence of the relationships between these taxa varied depending on which model was
used to deal with polymorphisms. A regular analysis in PAUP and the DPCWH method
have C. leanira in a basal taxon to the other three, with fulvia and cyneas coming out as
sister taxa in the regular PAUP analysis and cyneas and cynisca coming out as sister taxa
with the DPCWH model. In contrast, C. leanira and C. fulvia come out as sister taxa
with the distinct state model. None of the results are compatible with Higgins (1981)
classification of these taxa. I consider the results of these analyses to be inconclusive as
to the relationships between these four taxa, because (1)1 had very few specimens of C.
cynisca available for study, so results dependent on characters known to vary within a
taxon could be arguably suspect for that reason alone, and (2) as I will argue in a
subsequent section, I am unsatisfied with any of the options available for analyzing
polymorphic characters with currently available phylogenetic software.


26
Gnathotrusia mundina, two individuals of Atlantea tulita, and three or more individuals
of the remaining taxa. The vesica was everted, or partially everted, for at least one male
specimen of every taxon in the data matrix except for Janatella leucodesma and Ortilia
liriope.
Since 1 found very little intraspecific variation in the samples dissected other than
for the exact shape of the saccus for some taxa, I find it very unlikely that dissecting
additional individuals of the taxa examined would result in changing character coding
schemes for the characters coded. Fewer female specimens were dissected relative to
males, and for taxa other than Chlosyne, Thessalia, and Charidrycis, in most cases I
dissected only one or two females per taxon.
For the genera Chlosyne, Thessalia, Charidryas, and Anemaca, all 41 Operational
Taxonomic Units (OTUs) I recognize have the same character state for each character in
the Melitaeini data matrix, with very few exceptions noted under the appropriate
characters. Consequently, I entered these taxa into the data matrix collectively as
Chlosyne A and Chlosyne B. This is preferable to entering these taxa into the matrix
collectively as one taxon, because a single collective OTU would not permit the analysis
to show if there is evidence for the monophyly of these taxa (if two or more taxa are
coded with identical states for all characters, that by itself does not imply monophyly).
Preparation of Melitaeini Genitalia for Character Analysis
I studied genitalic characters from genitalia with their natural shape intact (as
opposed to putting genitalia on slides as was done by Higgins (1960) and Higgins (1981),
which involves destroying the three dimensional structure of Melitaeinine genitalia).
Genitalia were dissected using standard techniques of removing a specimens abdomen


694
small cream colored patches. I find that all these characters can vary within individuals
from the same geographic area, and the variation between extreme forms occurs along a
continuum; hence, I recognize no subspecies. Higgins (1960) suggested the presence of
cream colored patches on the ventral hindwing may be a subspecies difference for
specimens from the southern part of the range; however, I have seen specimens with and
without these markings from both Mexico and Panama.
Range: I have seen material from southern Mexico (states of Chiapas, Guerrero,
Veracruz, Oaxaca, Nayarit, and Quintana Roo), Guatemala, Honduras, Costa Rica,
Panama, Venezuela, and two specimens from Colombia lacking locality data. Data from
the NMNH and FSCA appears in Table 12.
Species Delimitation: Chlosyne hippodrome has unique male genitalia with distinct gaps
in variation compared to any other Chlosyne. Also, C. hippodrome has no sister species.
There are also gaps in wing pattern characters between C. hippodrome and any other
Chlosyne.
Remarks: Higgins (1960) reported the name Papilio hyperia Fabricius, 1973, Ent. Syst.
3:119 to be invalid, and proposed form fabricii as a replacement for specimens which
show yellow markings on the ventral basal hindwing, despite reporting that intermediate
specimens were examined between this condition and lacking any yellow at the ventral
hindwing base.
Chlosyne narva (Fabricius)
Figures 152,172,174,203,233,252 & 259:a-b.


474
orange band between the position of the marginal and submarginal bands (this area is
black in M. elva and M. dymas), a number of other character states listed in Table 5.
The genitalia of both sexes of M. eleda are unique and very different from any
taxon with a similar wing pattern. M. eleda has a diamond shaped juxta plateau ( Figure
108) with no posterior juxta process in contrast to the triangular plateau terminating in a
posterior process of M. dymas (Figure 122). The posterior valve projection points inward
and slightly dorsally in M. eleda (Figure 115) compared with strongly dorsally and
outward at the tip in M. dymas (Figure 119). The female genitalia of M. eleda have the
ridge on the lamella antevaginallis of nearly uniform width (Figure 280), while in M.
dymas it is greatly produced at the sides and reduced in the middle (Figure 281). The
female genitalia of Chlosyne acastus and its relatives lack a raised ridge on the lamella
antevaginallis, and the posterior valve projection in males is orientated dorsally at the tip
(Figure 210).
Further Description: Both orange and black scales and hairs on the labial palpi. Dorsally
palpi with predominately black scales and hairs with scattered white and orange scales
and hairs; laterally on the outer side with the white band not extending distal of the eye
and with orange and less black around the white band; ventrally orange and a thin sparse
row of black hairs on the inner ventral side and white at the base; laterally on the inner
side with a white band not reaching the terminal segment and predominately orange
scales and hairs ventrally and at base with predominately black scales and hairs dorsally
and at the tip. Vertex with black and orange scales, and white scales either scattered or
occasionally coalesced into a patch not reaching the posterior margin of the vertex.
Sutures at the lateral edges of the vertex with black, orange, and white scales. Tuft of


434
CALEOS Model
A B C D
0 O&l 1 1
A C B D
0 1 O&l 1
Distinct State Model
A B C D
0 0 11
0 111
A C B D
0 10 1
0 111
Figure 338: Behavior of different models for polymorphic characters on three tree topologies for
scenario 2, with the polymorphic character mapped on the trees. A-B: CALEOS Model. C-D: Distinct
State Model. E-F: DPCWH Model. G: Data matrix for scenario 2. The polymorphic character
requires two steps on trees A-B in PAUP 4.0b4a


Table 1-Continued
134
Characters
Taxa
41
42 43 44 45
46 47 48 49 50
51
52
53 54 55
56 57 58
59 60
Eurodryas aurinia
1
0
2
1
1
1
1
1
1
6
1
0
0
0
1
1
5
2
1
1
Eurodryas desfontaini
1
0
2
1
1
1
1
1
1
7
1
0
0
0
1
1
5
2
1
1
*Hypodryas maturna
1
0
2
1
1
1
1
1
1 5
1
0
0
J)_
1
1
2
2
0
1
Hypodryas intermedia
1
0
2
1
1
1
1
1
1
5
1
0
0
0
1?
1
2
2
0
1
Hypodryas gillettii
1
0
2
1
1
1
1
1
1
3
1
nr
0
0
1
1
2
2|
0
1
Hypodryas cynthia
1
0
2
1
1
1
1
1
1
5
1
0
0
0
1
1
2
2
0
1
Hypodryas iduna
1
0
2
1
1
1
1
1
1
5
1
0
0
0
1
1
2
2
0
1
Occidryas anicia
1
0
2
1
1
1
1
1
1
1
1
0
0
0
1
2
2
2
0
1
Occidryas chalcedona
1
0
2
1
1
1
1
1
1
1
1
0
0
0
1
2
2
2
0
1
Occidryas colon
1
0
2
1
1
1
1
1
1
1
1
0
0
fo
1
2
2
2
0
1
*Euphydryas phaeton
1
0
2
1
1
1
1
1
1
4
2
0
0
0
1
2
2
2
0
1
Occidryas editha
1
0
2
1
1
1
1
1
1
2
2
0
0
0
1
2
2
2
0
1
Out Group Taxa
Colobura dirce
0
0
1
1
0
1
1
1
7
0
0
0
0
0
3
3
?
?
?
1
Kallima paralekta
0
0
1
1
0
1
1
1
8
0
0
0
0
0
2
1
?
?
?
1
Anadia jatrophe
0
0
1
3
0
1
1
1
6
0
0
0
[ 0
0
2
1
?
?
?
1
Anadia amathea
0
0
1
3
0
1
1
1
6
0
0
0
0
0
2
1
?
?
?
1
Anadia fatima
0
0
1
3
0
1
1
1
6
0
0
0
0
0
2
1
?
?
?
1
Anadia chrysopelea
0
0
1
3
0
1
1
1
6
0
0
0
0
0
2
?
?
?
?
1
Nymphalis polychloros
0
0
1
1
0
1
1
1
b
0
0
0
0
0
2
1
?
?
?
1
Hypanartia lethe
0
0
1
1
0
1
1
1
0
0
0
0
0
0
2
3
?
?
?
2
Vanessa atlanta
0
0
1
1
0
1
1
1
c
0
0
0
0
0
4
1
?
?
?
2
Cynthia cardui
0
0
1
1
0
1
1
1
c
0
0
0
0
0
4
1
?
?
?
2
Hypolimnas pandarus
0
0
1
1
0
1
1
1
9
0
0
0
0
0
2
3
?
?
?
2
Amnosia decora
0
0
1
1
0
1
1
1
0
0
0
0
0
0
2
4
tt
?
?
2
Araschnia levana
0
0
1
1
0
1
1
1
a
0
0
0
0
0
2
1
?
?
?
2
Araschnia prorsa
oj
0
1
1
0
1
1
1
a
0
0
0
0
0
2
1
?
?
?
2
Agais udicae
ot
0
1
1
0
1
1
1
c
0
0
0
0
0
2
2
?
?
?
2
Precis octavia
0
0
1
1
0
1
1
1
5
0
0
0
0
0
2
1
?
?
?
1
Catacroptera cloanthe
0
0
1
3
0
1
1
1
0
0
0
0
0
0
2
2
?
?
?
2
Inachis io
0
0
1
1
0
1
1
1
d
0
0
0
0
0
2
1
?
?
?
2
Junonia coenia
0
0
1
1
0
1
1
1
0
0
0
0
0
0
2
1
?
? I
?
1
Polygonia c-aureum
0
0
1
1
0
1
1
1
c
0
0
0
0
0
2
1
?
?
?
2
Metamorpha elissa
0
0
1
1
0
1
1
1
6
0
0
0
0
0
2
1
?
?
?
2
Siproeta epaphus
0
0
1
1
0
1
1
1
0
0
0
0
0
0
2
1
?
?
?
1
Yoma sabina
0
0
1
1
0
1
1
1
0
0
0
0
0
0
2
1
?
?
?
2
Doleschallia bisaltide
0
0
1
1
1
1
1
1
0
0
0
0
0
0
2
1
?
?
?
1
Salamis augustina
0
0
1
1
0
1
1
1
7
0
0
0
0
0
2
1
?
?
?
2
Napeocles jucunda
tt
0
1
3
0
1
1
1
c
0
0
0
0
0
2
1
?
TT
?
2
Rhinopalpa polynice
0
0
1
1
0
1
1
1
6
0
0
0
0
0
2
1
?
?
?
1
Vanessula milca
0
0
1
3
0
1
1
1
a
0
0
0
0
0
2
1
?
?
?
1
Antanartia delius
0
0
1
1
0
1
1
1
c
0
0
0
0
0
2
1
?
?
?
2
Kaniska canace
0
0
1
1
0
1
1
1
c
0
0
0
0
0
2
1
?
?
?
2
Cumulative Out Group
0
0
1
1,c 0,1
1
1
1
0,c
0
0
0
0
0
2
U.c
?
?
?
1,2


174
42.5X
Figures 158-161. Male genitalia of Chlosyne sp. In dorsal aspect. 158: C. nycteis.
159: C. gorgone. 160: C. harrissii. 161: C. kendallorum.


239
furthermore such a state would not matter since it would be autapomorphic, occurring
only in a taxon divergent from taxa with state 1. Also, occasional individuals of the
western Charidryas group (C. hojfmani, C. palla, C. acastus and taxa with identical
genitalia) were found to have a spine on one valve proximal to the bend in the posterior
valve process. However, since this feature was found so infrequently, it is conceivable it
could occur in a taxa where it was not found, and it was not coded as a separate character.
28. Posterior dorsal border of the valve area with many hair-like setae:
0=Similar to the remainder of the valve with many hair-like setae.
l=With a thickened, heavily sclerotized U-shaped area at the dorsal edge (Figure
205).
All Melitaeini have an opening on the ventral side of each valve, formed by where the
plate comprising the valve twists around and overlaps with itself.
29. The form of the above ventral valve opening:
0=A distinct opening (Figures 80,94,108,110,112,122 & 134-155).
1=A closed slit (Figures 87,89 & 106).
30. The anterior extent of the ventral valve opening in the preceding character:
0=Termineates well posterior to the vinculum (Figures
87,89,94,99,106,108,110,112,122 & 134-155).
l=Terminates anterior of the vinculum (Figure 80).
All Melitaeini have a patch of setae on the inner wall of each valve in the vicinity of the
midpoint, visible in ventral aspect. In most taxa, the setae and their sockets are visible in
ventral view, but a couple taxa have the valvae folded such that the sockets are
concealed.


346
gaudealis (Bates)
a. gaudealis
b. wellingi Miller & Rotger
janais (Drury)
rosita Hall
a. rosita
b. browni Bauer
c. montana Hall
d. riobalensis Bauer
e. mazarum Miller & Rotger
Phylogeneticallv Invalid or Unsupported Generic Concepts
Several genera of Chlosyniti (which have been used by various authors since the
last revision of Chlosyne by Higgins (I960)) are not in their current usage compatible
with a natural classification scheme. Such genera are not monophyletic (Texola and
possibly Charidryas) or they make a different genus paraphyletic (Dymasia, Thessalia,
and Anemaca). I use the phylogenetic hypothesis which I have derived to convert
existing classifications into a classification scheme based on the criteria of monophyly
and stability with the guidelines I specified in the preceding chapter.
Texola Higgins, Dymasia Higgins, and Microtia Bates:
The phylogenetic analyses in the preceding chapter indicated that Higgins (1960
& 1981) concept of Texola (including eleda, coracara, and anomalus) was invalid,
because the monotypic Dymasia and Microtia come out in the middle of Texola. All six
parsimony analyses with the entire data set based on the Chlosyniti/Poladryiti data matrix
confirm this. Higgins (1960) basis for erecting the genus Dymasia was these
differences [between eleda and autapomorphic characteristics of dymas] are so marked
that 1 have found it impossible to write a generic synopsis to include both dymas and
eleda (Higgins 1960). Higgins (1960) argued in favor of a monotypic concept of


PHYLOGENETIC STUDIES OF THE MELITAEINI (LEPIDOPTERA:
NYMPHALIDAE: NYMPHALINAE) AND A REVISION OF THE GENUS
CHLOSYNE BUTLER
By
HUGO L. KONS, JR.
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE
UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE
REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
2000


519
Gatrelle's (1998) Figures of "subspecies" gorgone appear to me to be well within the
normal range of variation of C. gorgone throughout its range and exhibit no gaps in the
range of variation for any character. Consequently, while Gatrelle (1998) stated
"gorgone varies slightly but consistently from carlotta in phenotype" and "These
differences alone are enough to validate subspecific status", I failed to find any evidence
consistent with this claim, including from the characters mentioned by Gatrelle and the
specimens he illustrates. Consequently, I conclude there is no evidence of morphological
discontinuity or even clinal variation between different parts of the range of C. gorgone,
and consequently there is no basis for recognizing subspecies on morphological grounds,
regardless of one's subspecies concept.
Gatrelle (1998) also noted that coastal South Carolina and costal Georgia
populations are univoltine, while he claimed cariota was multivoltine throughout its
range, and reported this as the strongest argument for recognition of two subspecies. Yet
Gatrelle (1998) presents no data to contradict Scott's (1986) report that univoltine
populations of G. gorgone occur in the Colorado mountains and Canada, although
Catling and Layberry (1988) noted C. gorgone had three broods (always?) in eastern
Ontario. The number of broods appears to vary between seasons within some
populations of C. gorgone, and I have noted at Mosquito Hill Nature Center in
Outagamie County, Wisconsin, some years there were two disjunct flights while during
other years there were three (perhaps in other parts of its range there is variation between
one and two broods between years?). In my view collecting was sufficiently frequent and
intense to argue against the possibility that some years the third brood was not detected. I
view the number of broods as a character which commonly varies within a taxon in


507
hypothesis ismeria is harrisii can not be validated, but it can be preferred since it requires
fewer ad hoc explanations. Consequently, application of the name ismeria approaching
any reasonable degree of confidence is impossible and always will be, assuming no type
specimens of ismeria are ever located. However, there is in fact a very good reason for
assigning ismeria to harrisii instead of nycteis, an action that represents at least as good
of a hypothesis as any alternative. According to the most recent version of the Code for
Zoological Nomenclature which came out after Gatrelle (1998), if an older name is found
for a name currently in use, and the name currently in use appears in over 40
publications, the name currently in use is retained (ICZN 1999). C. nycteis and C.
harrisii have both been in universal and widespread use for decades and appeared in
more than 40 publications, but ismeria precedes both names. Application of the name
ismeria to harrisii results in the retention of the name harrisii and thus avoids the loss of
nomenclatural stability and confusion that would result from sinking a long standing,
widely used name (as proposed by Gatrelle (1998) with regard to changing the name of
C. nycteis to C. ismeria) in favor of a blatant nomen incognitum whose original
application will never be known.
Chlosyne kendallorum Opler
Figures 137,161,179,198,216,237,289 & 292.
Chlosyne kendallorum Opler, 1999. Holarctic Lepidoptera. 6(1): 23-24. Type Locality:
Mexico: Nuevo Leon: 40km WSW of Cola de Caballo, 13 May 1978. Holotype:
NMNH (Opler 1999).


24
are made regarding character state order, (2) character coding schemes which weight the
same change more than once are avoided, (3) coding schemes which may produce
unwarranted bias regarding character state polarity are avoided, (4) a continuum of
variation is not arbitrarily divided up into a series of discrete states, (5) characters which
seem to be completely dependent (such as two separate wing patches always being the
same color-more relevant to the following chapter) are coded only once, (6) for
characters involving further division of a character state into additional states which do
not apply to taxa lacking the state to be divided, "?" coding is avoided if but only if this
can be accomplished without weighting an identical step twice, and (7) characters which
can be coded into discrete states are not deleted because a priori I feel they may be
homoplastic.
In Group Taxa Examined
Taxa examined and coded for characters are listed in Table 1. All described
Melitaeini taxa other than Phycioditi and Eurasian Melitaeiti were included with the
following exceptions for which I did not have material available for study: Eurodryas
alexandria (Staudinger), Eurodryas orientalis (Herrich-Schafter), Gnathotrusia steinii
(Dewitz), and Atlantea cryptadia. Also, I did not dissect the few specimens I examined
of Adantea perezi (Herrich-Schaffer); however, judging from figures 11-13 (pg. 391) in
Higgins (1960), I suspect there would be very few if any differences in character coding
between this taxon and the other Atlantea examined. Several taxa had to be coded for
males only, since I had no females available for study. These taxa are Texola anomalus
(Godman & Salvin), Higginsius fasciatus (Hopffer), Gnathotriche sodialis Staudinger,
and Gnathotrusia mundina (Druce).


513
proximal band of the central symmetry system with orange scaling inside the same color
orange as that inside the discal cell. Dorsal forewing discal cell with a single orange
patch between the proximal band of the central symmetry system and the discal spot.
Basal band of basal symmetry system forming a basal patch in the discal cell with diffuse
orange scaling of identical color to the orange inside the discal spot. Dorsal forewing cell
CuA2 with the distal element of the basal symmetry system single and forming a narrow,
elongate patch with diffuse orange scaling inside extending from the basal part of the cell
almost to the origin of vein CuA2. Dorsal hindwing postmedian band present and in the
form of a series of orange patches in cells CuA2-R5, ventrally patches in cells CuA2-
Sc+Rl with variably sized brown patches in CuA2-Ml with a continuation (variable in
extent) of narrower patches in R5 and Sc+Rl along the basal side of the postmedian band
with cream background color distally (Sc+Rl may contain a larger brown patch).
Forewing postmedian band of orange patches with cream colored postmedian dots inside.
Hindwing postmedian dots present within the postmedian band patches in cells CuA2-Ml
and composed of black scales and in some cells cream colored scales in the center, with
the cream scaling often more extensive ventrally. The dorsal postmedian dot in cell M3
almost always has cream colored scales in the center and sometimes lacks a complete
border of black scales (no other postmedian dots lack a complete black border); the dorsal
postmedian dot in cell CuAl usually has cream colored scales also, and the dorsal dots in
cells CuA2 and M2 occasionally do, but more often they are all black; the Ml dot is solid
black. A ring of light colored scaling (orange or cream colored) contrasting with the
ventral hindwing postmedian band encircles some or all of the postmedian dots, and in a
few specimens the orange rings are so expanded that the feature approaches the


769
PLATE C


481
hindwing with no white sections between the veins. The usual narrow band of orange
scales along the basal forewing costal margin is present.
Dorsal forewing black and orange, with the size of the orange markings and
relative amounts of black background variable. One orange band partially or completely
encloses the forewing apex, and may contain a cream area inside of the orange. An
additional orange marking occurs on the dorsal forewing's posterior medial border,
sometimes extending to the base of the wing.
The dorsal hindwing has a black background with a transverse orange median
band transversing the entire wing from cell 3A to Sc+Rl and sometimes extending to the
basal area, and sometimes with a cream band inside of the orange. Postmedian dots are
absent dorsally and ventrally on all wings.
Ventrally a short orange patch is present in the basal forewing costal cell,
representing a homoplastic character state strikingly similar to the condition of many
derived Chlosyne. The apical forewing band is somewhat larger and sometimes longer
(up to twice as long in the extreme case) ventrally, and is predominately cream with
narrow bands of orange on the basal and distal sides. The other orange forewing patch is
the same size or smaller ventrally, apparently never extending all the way to the base of
the wing even if it did so dorsally.
The ventral hindwing median band is the same size or narrower ventrally, and
always occurs as a narrow transverse band never expanded as a patch into the basal part
of the wing. The color of the band is pale cream contrasting with its dorsal color and the
orange or darker cream ventral forewing markings, although narrow bands of orange


Table 12 Continued
748
NMNH=Nat Mus. of Natural Hist.-Smithsoman Inst. FSCA=FL St. Collection of ArthropoOs, AM=Allyn Mus.. PC=Personal Collection: @=Judged bv author as of dubious authenticity
County (USA)
Locality (USA)
Taxon
1 Country
State/Prov.
Locality (non US)
_
1
§
I
6
j
Date
Collector
Collect ior
Chlosyne lacinia Nicaragua ? Santa Maria de Ostuma
Nov 1959
N L H Krauss
NMNH
,
Note: NMNH specimens from Costa Rica are forms quehtala. lacinia, and intermediates between these forms
Chiosyne lacinia
Costa Rica
Alajuela
San Mateo. 1-2000 ft
Sep.
W Schaus
NMNH
2
Chiosyne lacinia
Costa Rica
Majuela
San Mateo. 1-2000 ft
Nov 25-Dec 12'06
W Schaus NMNH
2
Chlosyne lacinia
Costa Rica
Cart ago
Juan Vinas. 2500-3500 ft
Nov. '06
W Schaus NMNH
3
Chlosyne lacinia

Costa Rica
Cartago
Tuis
June
Schaus & Barnes NMNH
1
Chlosyne lacinia
Costa Rica
Cartago
Turnaba
20 May 1951
O L Cartwright NMNH
t
Chlosyne lacinia
L
Costa Rica
Cartago
Turrialba
26 May 1951
O L Cartwright NMNH
1
Chlosyne lacinia
Costa Rica
Cartago
Turrialba
2-5X1 1967
E L Todd
NMNH
Chlosyne lacinia
Costa Rica
Guanacaste
5 km NW Canas Hac La Pacifica. 50m
XI-5-71
Opler
NMNH
2
Chlosyne lacinia
Q
Costa Rica Guanacaste
Nr. Playa del Cocos
20-VI-1976
J K Bale lunas
FSCA
1
Chlosyne lacinia
Costa Rica Heredia
San Antonio de Belen, 900m
17-VIII 1977
GB Small
NMNH
2
Chlosyne lacinia
UQ
Costa Rica Puntarenas
Atenas 20 miles W San Jose
7-VII 1971
H L King
FSCA
2
Chlosyne lacinia
Costa Rica Puntarenas
Esparta. 900 ft
Oct. "06
W Schaus NMNH
3
Chlosyne lacinia
Costa Rica Puntarenas Palmar Norte. 400'
22 June 1970
NMNH
1
Chlosyne lacinia
Costa Rica Puntarenas Palmar Norte
23 June 1970
James H Baker NMNH
1
Chlosyne lacinia
UQ
Costa Rica Puntarenas Palmar None
5-VI-1971
H L King FSCA
,
Chlosyne lacinia
UQ
Costa Rica Puntarenas
Palmar Norte
10-VI-1971
H L King FSCA
,
Chlosyne lacinia
UQ
Costa Rica Puntarenas
17 miles E Palmar Norte
16-VI-1972
H L King FSCA
,
Chlosyne lacinia
O
COSta RiCa Puntarenas I7 miles E Palmar Norte
16-VI-1972
H L Kmg I FSCA
1
Chlosyne lacinia
Q
Costa Rica Puntarenas
Piedras Blancas
2.26.71
FSCA
1
Chlosyne lacinia
UQ
Costa Rica Puntarenas
20 km E Puntarenas
1/04/74
EC Olson
FSCA
Chlosyne lacinia
L
Costa Rica Puntarenas
Rio Palmare General. Ferry Crossing near Brujo, Pan Amor. Hwy X San Vito Rd.. 500'
10/2/86
T C Emmel
FSCA
2
Chlosyne lacinia
UQ
Costa Rica Puntarenas
20 miles W San Jose Arenas
7-VII-1971
H L King
FSCA
1
Chlosyne lacinia
Costa Rica Puntarenas San Vito
4-1-60
FSCA
I
Chlosyne lacinia
Q
Costa Rica Puntarenas San Vito River
12 March 1968
T C Emmel
FSCA
1
Chlosyne lacinia
Costa Rica Puntarenas villa Neiiy, 550m
15 VI1976
Gordon B Small
NMNH
1
Chlosyne lacinia
Costa Rica San Jose Puriscai Mts.
Sep.
W Schaus
NMNH

Chlosyne lacinia
Costa Rica San Jose sta Ana
8-VIII 1980
GB Small
NMNH
,
Chlosyne lacinia
Q.LC
Costa Rica San Jose
2-X-1966
DH Habeck
FSCA
12
Chlosyne lacinia
L
Costa Rica San Jose
2-X-1966
DH Habeck
FSCA
2
Chlosyne lacinia
Costa Rica San Jose
EdwTOwen
NMNH
5
Chlosyne lacinia
Costa Rica San Jose
NMNH
,
Chlosyne lacinia
L
Costa Rica
?
?
Chilamate-Finca de Selva 10 km from Puerto Vieja on Parapiqui R
12/13/85
EC Olson
FSCA
1
Chlosyne lacinia
L Costa Rica
Chilamate-Finca de Selva 10 km from Puerto Vieja on Parapiqui R.
12/13/86
EC CHson
FSCA
1
Chlosyne lacinia
l Costa Rica
1
Chilamate-Finca de Selva 10 km from Puerto Vieja on Parapiqui R.
12/15/86
EC Olson
FSCA
i"
Chlosyne lacinia
l
L
Costa Rica
Chilamate-Finca de Selva 10 km from Puerto Vieja on Parapiqui R.
12/20/86
EC Olson
FSCA
;
Chlosyne lacinia
Costa Rica ? Chilamate-Finca de Selva 10 km from Puerto Vieja on Parapiqui R.
12/25/86
EC Olson
FSCA
Chlosyne lacinia
L
COSta Rica I? Chilamate-Finca la Selva Verde
24 Sept. 1986
D Parkinson
FSCA
Chlosyne lacinia
Costa Rica ? S Domingo
Sept
W Schaus
NMNH
3
Chlosyne lacinia
Costa Rica ?
EdwTOwen
NMNH
5
Chlosyne lacinia
Costa Rica ?
B Neumogen
NMNH

Chlosyne lacinia
Panama Canal Zone CocoH
west Panama (Rio Armila lo Oanan) include these forms plus form saundersom and lacinia saundersom and
late Oct 1962
quehlale seundereow ir.ermd.wej
NMNH
Chlosyne lacinia
Panama Canal Zone Coco Solo
20 March 1944
W H Wagner Jr
NMNH
2
Chlosyne lacinia
Panama Canal Zone Coco Sdo
21 March 1944
W H Wagner Jr
NMNH
Chlosyne lacinia
Panama Canal Zone Farfan
XII-25-62
NMNH
,
Chlosyne lacinia
Panama Canal Zone Fort shermat
XI-4-62
NMNH
,
Chlosyne lacinia
Panama Canal Zone Fort shermat
XI-10-62
NMNH
2
Chlosyne lacinia
Panama Canal Zone La Boca
Nov 5-07
T H Hallman
NMNH
,
Chlosyne lacinia
uq Panama Canal Zone La Boca
Jan 22: 08
T H Hallinan
NMNH
,
Chlosyne lacinia
o Panama Canal Zone La Boca
Jan 22: 08
T H Hilinai
NMNH
Chlosyne lacinia
UQ
Panama Canal Zone Madden r. f
2-11-68
FSCA
1
Chlosyne lacinia
UQ
Panama Canal Zone Panana Canal
10 Oct 67
H O Hilton
FSCA
Chlosyne lacinia
Panama Canal Zone Paraiso
20.II
NMNH
,
Chlosyne lacinia
Q
Panama Canal Zone Summit
11 XII 1978
R Robbins
NMNH

Chlosyne lacinia
Q
Panama Canal Zone Summit
3 VI 1979
R Robbins
NMNH
t
Chlosyne lacinia
L.Q
Panama Canal Zone summit
14 VII 1979
R Robbins
NMNH
,,
Chlosyne lacinia
Q
Panama Canal Zone Summit
16 VII 1979
R Robbins
NMNH
2
Chlosyne lacinia
L.Q
Panama Canal Zone Summit
20 VII 1979
R Robbins
NMNH
,2
Chlosyne lacinia
Q
Panama Canal Zone summit
6 VIII 1979 R Robbins
NMNH
2
Chlosyne lacinia
Q
Panama Canal Zone
25/67 H O Hilton
Chlosyne lacinia
Panama Chiriqui
EdwTOwen
NMNH
Chlosyne lacinia
L
Panama Colon Pina
2-13-64
FSCA
1
Chlosyne lacinia
Q.L/C
Panama Colon Pina
4-IV-1970 H L King
FSCA
Chlosyne lacinia
L.L/Q
Panama Colon Pina. 200m
21-III 1971 H L Kmg
FSCA
Chlosyne lacinia
L