Citation
Protein variation in Gomphus (Odonata: Gomphidae)

Material Information

Title:
Protein variation in Gomphus (Odonata: Gomphidae)
Creator:
Knopf, Kenneth William, 1950-
Publication Date:
Language:
English
Physical Description:
vi, 107 leaves : ill. ; 28 cm.

Subjects

Subjects / Keywords:
Alleles ( jstor )
Dehydrogenases ( jstor )
Enzymes ( jstor )
Gels ( jstor )
Genetic distance ( jstor )
Genetic loci ( jstor )
Genetic mutation ( jstor )
Phylogeny ( jstor )
Population estimates ( jstor )
Species ( jstor )
Dissertations, Academic -- Entomology and Nematology -- UF
Dragonflies -- Genetics ( lcsh )
Entomology and Nematology thesis Ph. D
Gomphus ( lcsh )
Odonata -- Classification ( lcsh )
City of Gainesville ( local )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis--University of Florida.
Bibliography:
Bibliography: leaves 103-106.
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Kenneth William Knopf.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
This item is presumed in the public domain according to the terms of the Retrospective Dissertation Scanning (RDS) policy, which may be viewed at http://ufdc.ufl.edu/AA00007596/00001. The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact the RDS coordinator(ufdissertations@uflib.ufl.edu) with any additional information they can provide.
Resource Identifier:
020620341 ( ALEPH )
03983201 ( OCLC )

Downloads

This item has the following downloads:


Full Text














PROTEIN VARIATION IN GOMPHUS (ODONATA: GOMPHIDAE)















By

KENNETH WILLIAM KNOPF


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












UNIVERSITY OF FLORIDA


1977
















ACKNOWLEDGEMENTS


I sincerely thank the members of my graduate committee, Dr. Dale H. Habeck, Chairman, Dr. Minter J. Westfall, Jr., Dr. Lewis Berner, and Dr. Milton Huettel for their assistance and guidance during this study. I give special thanks to Dr. Habeck who has provided excellent facilities and friendly encouragement throughout my graduate study.

Sincere appreciation is extended to Dr. Huettel for his valuable advice and the use of his laboratory which has made this study possible. I thank Ms. Candy Woodburn and Ms. Winifred Gaddis, of Dr. Huettel's laboratory, who were responsible for much of my instruction in the technical aspects of electrophoresis.

A special debt of gratitude is due Dr. Westfall, who was responsible for kindling my interest in Odonata.

The most important acknowledgement to be made is to my wife, Suanne, wixm I thank for providing technical, financial and emotional support.

I thank Dr. and Mrs. William L. Peters, Florida A & M University, Tallahassee, for their warm hospitality and for the use of the University facilities in the Blackwater River State Forest.

I gratefully acknowledge the assistance of Dr. Paul

Fuerst, of the University of Texas at Houston, with computer programming of allele frequency data.

i









I also express my appreciation to the Florida Department of Natural Resources for financial support during my graduate study.


iii
















TABLE OF CONTENTS

Page

ACKNOWLEDGEMENTS . . . . . . . . . . . . . . . . . . . ii

ABSTRACT . . . . . . . . . . . . . . . . . . . . . . . v

INTRODUCTION . . . . . . . . . . . . . . . . . . . . . 1

METHODS AND MATERIALS . . . . . . . . . . . . . . . . 5

Biochemical Background and Terminology. . . . . . 5
Overview of General Procedure . . . . . . . . . . 8
Species Studied . . . . . . . . . . . . . . . . . 9
Enzyme and Buffer Systems . . . . . . . . . . . . 10
Preparation of Gels . . . . . . . . . . . . . . . 13
Preparation of Specimens. . . . . . . . . . . . . 15
Running of Gels . . . . . . . . . . . . . . . . . 16
Staining, Fixation and Storage . . . . . . . . . 19
Scoring of Gels and Data Processing . . . . . . . 20

RESULTS . . . . . . . . . . . . . . . . . . . . . . . 23

DISCUSSION . . . . . . . . . . . . . . . . . . . . . . 66

Percent Polymorphism . . . . . . . . . . . . . . 66
Heterozygosity . . . . . . . . . . . . . . . . . 67
Genetic Distance and Proposed Phylogeny . . . . . 70
Neutralists Versus Selectionists . . . . . . . . 86
Evolutionary Clocks . . . . . . . . . . . . . . . 87

SUMMARY . . . . . . . . . . . . . . . . . . . . . . . 89

APPENDICES

A LOCALITY DATA . . . . . . . . . . . . . . . . . . 91

B BUFFER AND STAIN FORMULATIONS . . . . . . . . . . 97

LITERATURE CITED . . . . . . . . . . . . . . . . . . . 103

BIOGRAPHICAL SKETCH . . . . . . . . . . . . . . . . . 107


iv















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



PROTEIN VARIATION IN GOMPHUS (ODONATA: GOMPHIDAE) By

Kenneth William Knopf

December 1977

Chairman: Dale H. Habeck
Major Department: Entomology and Nematology

Protein variation in the dragonfly genus Gomphus was studied using starch gel electrophoresis. A phylogeny was proposed for 23 species using genetic distance estimates derived from analysis of 22 genetic loci. An additional 23 species of Odonata from a wide variety of families were analyzed for comparison with Gomphus.

Average heterozygosities (H) for 23 species of Gomphus ranged from 0.0000 - 0.0852, with a grand mean of 0.0221. Six species of Gomphus had no apparent variability. Nine species in Libellulidae had a mean H of 0.0491. These H values are much lower than most other insects that have been investigated.

The range of genetic distance levels for local populations was 0.0000 - 0.0191, for sibling species 0.0438 - 0.0868 and for genera 0.5023 - 1.783. In general the proposed phy-


v




vi


logeny agreed rather well with the existing classification of Odonata as established from conventional morphological studies. On the basis of genetic distance Arigomphus was raised from subgeneric to generic rank. Gomphus brimleyi Muttkowski, identical at all 22 loci with G. cavillaris Needham, was synonymized.


Chairman














INTRODUCTION


The status of the subgeneric rankings within the genus Gomphus have been in dispute for some time. Needham (1901) first organized the known North American species of Gomphus into four subgenera: Arigomphus, Gomphus, Gomphurus and Stylurus. Needham (1948) proposed that Arigomphus and Stylurus be elevated to generic status. In 1951, he introduced a fifth group, Hylogomphus, comprised of 4 species previously included in the subgenus Gomphus. In 1955, Needham and Westfall treated all five groups as subgenera. Some authors accepted the elevation of Stylurus even before Needham's 1948 paper. Williamson (1932) described 2 new species and placed them in the genus Stylurus. Walker (1958) followed the subgeneric rankings of Needham and Westfall (1955) but stated he had found Arigomphus, Stylurus and Gomphurus to be clearly defined groups based on study of the penis, vulvar lamina and hamules. Fraser (1957) recognized Stylurus and Gomphurus as valid generd but not Arigomphus.

The analysis of electrophoretically demonstratable protein variation has proven to be one of the most important tools of systematists and population geneticists in the study of evolution (Avise, 1974). Much of the pioneering application of this technique to questions of evolutionary


I





2


concern was conducted on Drosophila (Hubby and Lewontin, 1966, Ayala and Powell, 1972, and Ayala and Tracey, 1973). Since the late 1960's a wide variety of organisms other than Drosophila have been subjected to electrophoretic scrutiny. Examples of typical studies, although by no means an inclusive list, include: Gorman and Kim (197() proposed a detailed phylogeny of Anolis lizards based on 22 enzyme loci, Selander et al., (1971) studied the variation between subspecies of the mouse, Peromyscus polionotus, King and Wilson (1975) compared humans and chimpanzees and Gorman et al., (1976) analyzed variation in littoral gobies.

The basic hypothesis underlying taxonomic analysis of

protein variation is that over evolutionary time, due to the random nature of mutation, isolated populations will accumulate electrophoretically detectable changes in allele frequencies. Changes should accumulate at a rate proportional to the mutation rate. Thus,the longer the time since isolation, the greater the likelihood that populations can be differentiated by electrophoresis.

The only work reported on dragonflies is that of Anderson et al. (1970). Hemolymph samples from larvae of Anax junius, Aeshna umbrosa and Libellula pulchella were compared using acrylamide gel disc electrophoresis and an AmidoSchwarz general protein stain. They reported clear separation of all three species and postulated that the technique might be valuable in describing phylogeny.




3


There is considerable controversy concerning the mechanisms responsible for the maintenance of polymorphic variation in natural populations. Kojima and Tobari (1969) and Johnson (1974) proposed that natural selection is responsible for the observed variation. Kimura (1969), Ohta (1972) and Nei (1975) are the main proponents of the neutral mutation hypothesis which says most of the molecular polymorphisms are maintained by the random walk of frequencies of selectively neutral alleles. The implications of this controversy for measures of genetic distance and estimates of time since divergence will be discussed later.

The specific objectives of my research were:

1) To obtain heterozygosity estimates for members of the genus Gomphus and other selected odonates and compare them with each other and with published heterozygosity values for other organisms.

2) To construct a phylogeny of Gomphus from electrophoretic genetic distance data and compare that phylogeny with the existing classification derived from conventional morphological studies.

Special emphasis was placed on resolution of the classification of the groups included in Gomphus.

3) To serve as a supplementary procedural guide for electrophoresis of dragonflies.

The discussion will include suggestions for further

research on other groups of Odonata where analysis of protein variation might significantly improve classification.




4



Prospective researchers are encouraged to consult more extensive methods manuals such as Bush and Huettel (1972), Brewer (1970), Shaw and Koen (1968) and Harris and Hopkinson (1976).
















METHODS AND MATERIALS

Biochemical Background and Terminology


Proteins are made up of long chains of amino acids, of which there are about 20 commonly occurring varieties. The majority of amino acids are electrostatically neutral. The acidic amino acids with negative charges are aspartate and glutamate. The basic amino acids lysine, arginine and sometimes histidine are positively charged (Stryer, 1975). The primary structure of proteins consists of the sequence of the amino acids which compose it. Secondarystructure refers to patterns formed by hydrogen bonding between amino acid residues close to each other in the primary sequence. Tertiary structure refers to the complex folding of the chain to determine the final shape of the molecule. A protein is said to have a quaternary structure when two or more separate protein chains are closely associated. The associated subunits can be identical, products of different alleles at a polymorphic locus or products of different but related loci. The most widely known example of the third possibility is the hemoglobin molecule in man (Stryer, 1975). Hemoglobin has four separate chains and is called a tetramer; since its subunits are not identical it is further defined as a heterotetramer. Multimers (having more than one unit) with identical

5





6


subunits are prefixed with homo-, such as a homodimer consisting of two equal subunits. Proteins with only one peptide chain are called monomers.

Electrophoresis can be defined as the separation of

differently charged particles in an electric field. Stryer (1975), in elaborating on the determinants of electrophoretic mobility, states:


E (Z)
V =
f

where: V = Velocity of migration

E = Strength of electric field

Z = Net charge on the protein

f = frictional resistance

The frictional resistance is a function of the size

and shape of the molecule versus the pore size of the medium. The water soluble proteins in this study are almost all globular in shape. Since allelic differences coding for alteration of one of a few amino acids in the vast structure of the whole protein would not be expected to have much effect on the shape, the frictional resistance can be ignored. The strength of the electric field is roughly identical for all specimens in a single gel and therefore can also be ignored. The velocity can then be considered directly proportional to the charge on the molecule. A change of even a single charge unit can be detected. Often this may represent mutation of only a single amino acid. The substitution of





7


neutral valine for negatively charged glutamate at the 6th position in the beta chain of hemoglobin A yields hemoglobin S. Hemoglobin S is responsible for sickle cell anemia. These 2 proteins, differing in only one amino acid, can be separated by electrophoresis (Stryer, 1975).

Since most amino acids in a protein are not charged,

their mutation to another uncharged amino acid is not detectable unless the shape of the molecule is drastically altered. Nei (1975) estimates the probability of a mutation causing an electrophoretically detectable charge change at roughly 0.25 to 0.30 using both theoretical and empirical data. Therefore, assuming accumulation of mutations over time, the longer the time since divergence, the higher the likelihood populations will be separable by electrophoresis. This relationship holds until very large genetic distances are reached. At distances greater than 2 or 3 the probability increases that charge differences will be cancelled out by additional mutations. Therefore, the distance between distantly related organisms will tend to be underestimated. Because of this phenomenon, differences between taxa should be stressed more than similarities.

The terminology in this report will, for the most part, follow that of Prakash and Lewontin (1968). The term isozyme is used to refer to 2 or more enzymes that can be stained simultaneously using the same procedure for both. They presumably have identical biochemical roles in the organism. Allozymes, mobility variants within a single locus, will most




8


often just be called alleles. Allozymes are therefore also isozymes,but isozymes, if coded for by different genes, are not allozymes. King and Wilson (1975) have proposed calling allozymes "electromorphs" to emphasize that electrophoresis only reveals a portion of the genetic variation present. I have continued with the more widely used allelic terminology.


Overview of General Procedure


Specimens to be electrophoresed were homogenized in separate drops of buffer. Each crude homogenate was then soaked up on one or more small pieces of chromatography paper. A horizontal slit was cut in the precast starch gel and the specimen papers inserted. A direct current was passed through the gel perpendicular to the slit for the papers. Proteins having a net positive charge migrate toward the negative cathode and those having a net negative charge toward the positively charged anode. Most enzyme, proteins are in the latter category. The length of time the current is applied is dependent on the consistency of the gel, buffer systems employed and the migration rate of the enzyme being studied. The gel was then trimmed, removed from its mold and sliced into thin sheets. These were stained for any of a large variety of specific enzyme systems by use of the appropriate histochemical technique. Gel slices were incubated in the stain solution at 370C, with occasional agitation until the banding pattern developed.





9


The gels were then preserved in fixative and refrigerated. After being photographed they were evaluated and scored.


Species Studied


A list of dragonfly species studied is given in Table 1. Where more than one population was studied, it is indicated by assignment of a population number. The taxonomic interpretation is close to that of Needham and Westfall (1955). The non-gomphine species were included to help establish approximate genetic distance criteria for taxonomic levels within the order and to have a wider base of comparison for the analysis of variation within Gomphus. Appendix A contains collection data and localities for all species studied. All determinations were made by the author.


Enzyme and Buffer Systems


An annotated list of recipes for stains and buffers

used in this study, along with a list of chemical abbreviations, are included in Appendix B. Table 2 lists the 22 enzymes, and their abbreviations, that were actually scored and represent the basis for all further computations. Table

3 lists the additional enzymes that were tried during preliminary screening. These were not scored for any of a variety of reasons such as faintness, poor resolution, lack of repeatability, or total lack of staining. Just the head and thorax of each specimen were used, the abdomens were saved as voucher specimens. Some of the enzyme systems not






10


Table 1. List of dragonfly taxa.


GOMPHIDAE
Gomphus
Subgenus Arigomphus
furcifer Hagen pallidus Rambur villosipes Selys

Subgenus Gomphurus
dilatatus Rambur
lineatifrons Calvert

Subgenus Gomphus
australis Needham
brimleyi Muttkowski cavillaris Needham
diminutus Needham
exilis Selys
hodgesi Needham
lividus Selys
minutus Rambur spicatus Hagen

Subgenus Hylogomphus
abbreviatus Hagen apomyius Donnelly
brevis Hagen
parvidens Currie viridifrons Hine
sp. n.

Subgenus Stylurus
plagiatus Selys
potulentus Needham
townesi Gloyd

Progomphus
obscurus (Rambur)
sp. n.

AESHNIDAE Anax
junius (Drury)


AESHNIDAE continued Coryphaeschna
ingens (Rambur)

Epiaeschna
heros (Fabricius)

Gomphaeschna
antilope (Hagen) furcillata (Say)

MACROMIIDAE Didymops
floridensis Davis
transversa (Say)

CORDULIIDAE Epitheca
costalis (Selys)
cynosura (Say)
sepia (Gloyd)
stella (Williamson)

LIBELLULIDAE Ladona
deplanata (Rambur)

Libellula
auripennis Burmeister
flavida Rambur
incesta Hagen
needhami Westfall
semifasciata Burmeister

Orthemis
ferruginea (Fabricius)

Pachydiplax
longipennis (Burmeister)

Plathemis
lydia (Drury)


Basiaeschna
janata (Say)





11


Table 2. Names and abbreviations for the 22 enzymes scored
in this study.


Name


Acid Phosphatase-l Acid Phosphatase-2 Adenylate Kinase Alcohol Dehydrogenase Aldolase

Esterase-1

Esterase-2

General Protein Glutamate Dehydrogenase a-Glycerophosphate Dehydrogenase Hexanol Dehydrogenase Hexokinase

b-Hydroxybutyrate Dehydrogenase Isocitrate Dehydrogenase-l Isocitrate Dehydrogenase-2 Lactate Dehydrogenase Leucine Amino Peptidase Malic Enzyme Phosphoglucomutase Phosphoglucose Isomerase Tetrazolium Oxidase-l Tetrazolium Oxidase-2


Abbreviation


ACPH-1 ACPH-2

ADK ADH ALD EST-1 EST-2 GEN PRO

GDH GPD

HEX-DH HEX HBD

IDH-l IDH-2 LDH.

LAP ME PGM

PGI TO-1 TO-2





12


Table 3. Enzyme systems surveyed but not scored.



Enzyme Staining?


Aldehyde Oxidase no

Alkaline Phosphatase yes

Fumerase yes

Galactose Dehydrogenase no

Glucose-6-Phosphate Dehydrogenase yes

Glutamate-Oxaloacetate Transaminase no

Glyceraldehyde-3-Phosphate Dehydrogenase yes

Leucine Amino Peptidase (old) no

Malate Dehydrogenase yes

Monoamine Oxidase yes

Octanol Dehydrogenase yes

Pepsinogen yes

6-Phosphogluconate Dehydrogenase no

Sorbitol Dehydrogenase no

Succinate Dehydrogenase no

Triosephosphate Isomerase yes

Tyrosinase no

Xanthine Dehydrogenase no









resolved may have been specific to abdominal tissues. In future studies where voucher specimens are not necessary some of these enzyme systems might prove useful. To save time for any future studies that might try to improve on the resolution, I have indicated which systems stained even though they could not be scored.

Table 4 indicates the most common buffer-enzyme combinations employed, duration of the run, and the amperage or voltage to be maintained during the run.


Preparation of Gels

A mixture of 34.6 g Connaught hydrolysed potato starch (Fisher Scientific, New York, N.Y.), 11.4 g Electrostarch (Electrostarch Co., Madison, Wisconsin) and 20 g sucrose was dissolved in 400 ml of gel buffer in a 1000 ml flask. The quantities of starch and sucrose were empirically chosen. They can be varied along with the relative proportions of each in an attempt to maximize band resolution. The mixture was then heated on a stirring hot plate until it was uncomfortably hot to the touch. At this point the stir bar was removed and, with an asbestos glove, the flask was swirled vigorously over the open flame of a gas burner. As the gel heats it thickens and becomes translucent. Heating was continued until the gel just began to boil. A slight reduction in viscosity may be noticed before the boiling point is reached.

The gel was then degassed for about 45 seconds using a Venturi-type water aspirator. The aspirator was connected




14


Table 4. The most common enzyme-buffer combinations used
in this study with the duration of run and voltage
and amperage settings.


I. Electrode buffer - Borate
Gel buffer - Tris-HCl


Run DurationSettingsEnzymes-


1 hours 250 V ACPH ADH


TO

II. Electrode buffer - Borate
Gel buffer - Poulik


Run durationSettingsEnzymes-


Until 7.5 to 8.0 cm brown
zone migration 180 V PGI
HBD LAP


III. Electrode buffer - Histidine-8.0
Gel buffer - Histidine-8.0


Run durationSettingsEnzymes-


IV. Electrode buffer - Tris
Gel buffer - Poulik


Run durationSettingsEnzymes-


4 hours 50 mAmps
GPD ALD PGM
IDH ME

Citrate


3 hours 50 mAmps GDH
HEX-DH HEX


3 hours 50 mAmps EST
ADK


4 hours
250 V LDH
GEN PRO





15


directly to a 2 hole stopper, suction being applied by fitting the flask to the stopper and covering the second hole with a finger. The vacuum must be released by removing the finger from the hole rather than by shutting off the water. Safety glasses and a shield are a necessity since there is a real danger of implosions. The degassed gel was then poured quickly into the center of a 1.0 cm deep, 17.5 x 19.5 cm square,plexiglas mold. This same mold was used to hold the gel until it was sectioned and stained following the run. Small pieces of debris could only be removed from the gel while it was still molten.

After the gels cooled to room temperature (about 1 hour), they were wrapped in clear plastic food wrap to retard drying. Gels stored for a day at room temperature give the most consistent results. Gels were refrigerated for just one or two hours prior to specimen application.


Preparation of Specimens


Specimens to be electrophoresed were killed by freezing in a cooler of dry ice while on field trips. On return to the laboratory they were transferred to an ultra low temperature freezer for storage. Temperatures well below freezing are required to prevent enzyme degradation and denaturation.

The head and thorax of each frozen specimen was pulverized in a round-bottom,2 ml,plastic vial using a Delrin R plastic rod. The abdomen was retained for confirmation of species identification. This whole operation was conducted




16


on a bed of dry ice with prechilled tools. Special care was taken to avoid any heat transfer from the hands to the samples. All grinding rods were carefully wiped clean after each specimen was pulverized to prevent contamination. Since some of the enzymes to be surveyed may have been tissue specific, pulverization of specimens made a potentially more homogeneous sample. An additional advantage of pulverization is that only a small portion of the sample is required for a run; thus the same individual can be run repeatedly. This flexibility allows many side by side comparisons between taxa found, in hindsight, to be closely related.


Running of Gels


A clean, metal spatula, moistened with gel buffer,

served to pick up an adequate subsample of the frozen, pulverized head and thorax without thawing the sample. A single pulverized specimen of Gomphus provided enough material for 5 runs. The subsamples were deposited in the separate holes of a prechilled DelrinR plastic grinding block with a few drops of grinding buffer. A tray of crushed ice was used to keep the block chilled during the entire process. Individually assigned specimen numbers were recorded on protocol sheets for each grinding. The sample was ground (mixed) using a Delrin grinding rod chucked into a variable speed stirrer until a relatively smooth homogenate was obtained (about 10 to 20 seconds). After mixing was complete for all individuals to be run, separate rectangles of 1.0 x 0.4 cm






17


WhatmanR No. 1 chromotography paper were used to soak up the homogenate. Up to 5 pieces of paper were used in each hole.

A slit was cut in the chilled starch gel 5.5 cm from one end using a clean metal spatula and a ruler. The slit in the gel was eased open with the fingers and the lightly blotted sample papers inserted with a forceps. Twenty-four separate papers of 0.4 cm width could be put in each gel with approximately 0.3 cm between each paper and 1.5 cm on each end of the row to allow room for trimming the gel. When the slit was closed the gel was ready to be set up for the run.

Details of the construction of the buffer boxes, power supplies and safety boxes used are described by Bush and Huettel (1972). Each plexiglas buffer box holds approximately 450 ml. Buffer boxes containing pLatinum wire electrodes were placed in a modified plastic food crisper and the remainder of the food crisper packed with crushed ice. The gel tray was placed on top of the buffer boxes with the slit in the gel parallel to the buffer boxes. Sponge cloths served as wicks to bring the buffer and current up into the gel. The gel was covered with clear plastic food wrap from a line parallel to the slit and 3.5 cm from the cathodal end of the gel to beyond the anodal end of the gel. The sponge cloth for the cathodal end of the gel was drawn up to about 1 cm from the slit in the gel. Thus the cathodal sponge overlapped the clear plastic wrap by 1 cm. The food




18


wrap helped prevent drying and shrinkage during the run. The anodal sponge was brought up under the food wrap to a line 9.0 cm anodal of the slit. The gel and sponges were then covered with a glass plate that served both for electrical insulation (safety) and as a support for a pan of ice to keep the gel from overheating during the run. The power supplies were connected and set (Table 4) and the run began. Electrical resistance of the gels change during the course of the run so settings must be checked repeatedly. With the borate electrode - Poulik gel buffer system, a brown zone could be seen where the consistence of the gel changed, presumably as the buffer of the electrode passed through the gel. With this system the run was terminated after a satisfactory migration distance was obtained (usually 7.5 to 8.0 cm). On buffer systems without a visible front, the length of the run was determined empirically. Table 4 contains the durations of runs used with each specific buffer-stain system.

After turning off the power, the gel in its tray was removed from the apparatus and trimmed with a clean metal spatula and a straight edge. A centimeter of gel was trimmed off and discarded from each side as well as off the cathodal end of the gel. A cut across the gel was made 9.0 cm anodally from the slit and the portion that was under the anodal sponge discarded. The remainder was the larger, approximately 16.5 x 9.0 cm front or anodal slice, and the approximately 4.5 x 16.5 cm back or cathodal slice. Before they were removed from the gel tray they were marked for future orientation. A notch was taken off the left side of





19


each gel on the corner away from the slit. Most enzyme systems travel in the anodal direction. Only TO-1 and sometimes PGI were cathodal migrants in the dragonflies studied. Therefore, during the majority of the runs a back slice was not stained. However, during preliminary surveys, a back slice was always stained to ascertain which way the protein migrated.

The trimmed gel was placed on a plexiglas slicing tray. Air bubbles were cleared from beneath the gel by pressing gently on the top of the gel while observing from below. The gel slicer was a modified hacksaw that holds a thin stainless steel wire instead of a blade. The slicer was brought through the gel in one smooth, continuous motion with the wire held firmly down against the 2 mm high guide rails of the slicing tray. The thin slice was removed for staining. Three and sometimes 4 slices could be obtained from each gel. After each cut the wire was carefully wiped clean to prevent small bits of gel from hardening on the wire and resulting in the following slices being very rough surfaced.


Staining, Fixation and Storage


Stain solutions were prepared as close as possible to the time of their use, since they often contain unstable cofactors and enzymes. The dry ingredients were weighed out while the gels were running and stored in the refrigerator in small flasks. Liquid components were added at the




20


last possible minute. Gel slices were incubated (370C) according to the appropriate recipe (Appendix B) until the banding patterns developed. The staining solution was then discarded and about 75 ml of fixative added. Gels changed from translucent to opaque as fixation progressed. The following day gels were removed from the fixative, photographed and wrapped in clear plastic food wrap (Saran seems to work best). Wrapped gels may be stored for more than a year in a refrigerator. Scoring was done as soon as possible, since a few of the systems fade with storage.


Scoring of Gels and Data Processing


As is necessary in all electrophoretic studies where mating tests are impractical, the banding patterns on the gels were interpreted as direct gene products. In organisms such as Drosophila (Hubby and Lewontin, 1966), Gryllus (Harrison, 1977) and Rhagoletis (Berlocher, 1976) where

crossing is possible, most enzyme banding patterns tested by breeding experiments proved to be genetically based allelic variation. Dragonfly biochemistry is not likely to be grossly different from that of these organisms.

The allelic form most common in Gomphus abbreviatus

was taken as a standard reference allele. Faster and slower forms were compared by dividing their migration distances into that of the standard. By continually varying the order the specimens were put on the gel, side by side comparisons were made between most of the possible combinations of species studied.





21


To estimate the degree of genetic similarity between

populations the gene frequency data for each population were compared in a pair-wise manner to all other populations. The result is a matrix of numerical estimates of similarity for all possible combinations.

The measure of similarity selected was Nei's (1978)

estimate for the unbiased genetic distance since it compensates for a slight overestimate by the standard formula when applied to small sample sizes. A genetic distance (D) of zero indicates the proteins of that pair of populations were electrophoretically identical at all loci sampled. The estimate of the standard distance is given by D = -ln(Gxy//dxGy ). The average population gene identities G , Gy, and Gxy are usually replaced by the observed sample frequencies J, Jy, and J y,which are the averages of
2
Zxi2,~yi 2, and Yxiyi over all loci, where xi is the gene frequency of a particular allele in population x and y. is the frequency of the same allele in population y. To obtain the unbiased estimate of D for populations X and Y where the number of individuals, n and n can vary from locus to x y
locus, G and G are estimated by the average over all loci of (2nxyx 2 - 1)/(2nx - 1) and (2n yyi 2 - 1)/(2nv - 1), respectively.

Heterozygosity estimates were obtained with the formula
2
H = 1 - Ex2 , averaged over all loci, where xi is the frequency of the i-th allele.





22


Copies of the computer programused to analyze the allele frequency data in this study are available upon request from Dr. Masatoshi Nei, Center for Demographic and Population Genetics, University of Texas Health Science Center at Houston, P.O. Box 20334, Houston, Texas 77025. The programs have also been adapted to the IBM 360 computer at the University of Florida (NERDC) and are available through Dr. M. D. Huettel, Insect Attractants Behavior and Basic Biology Research Laboratory, USDA, Gainesville, Florida.
















RESULTS


Observed allele frequencies for Gomphus and Progomphus are presented in Tables 5-9. A total of 110 alleles were resolved in the 22 loci surveyed in Gomphus. With the addition of Progomphus and the rest of the species studied in other families of Odonata, the total number of separate alleles was 304.

Relative allele mobilities of Gomphus and Progomphus

are given in Figures 1-22. Of the 22 loci scored, 10 (45.5%) were monomorphic in all species of Gomphus. Four of those 10 (TO-l, GDII, ACPH-1 and ALD) exhibited a single monomorphic allele occurring in all species.

Percent of loci polymorphic and percent loci variable for Gomphus and Progomphus are presented in Table 10.

Table 11 presents heterozygosity estimates for all loci and the average heterozygosity (R) for each population of Gomphus and Progomphus. Average heterozygosities for Gomphus ranged from 0.0000 to 0.0852 and the mean for the 23 taxa was 0.0221. Six populations had no apparent variability. The heterozygosity data for the species of Odonata studied that were not in the family Gomphidae are summarized in Table 12. The non-gomphines were included just for comparison purposes and possible heuristic value. Many of 23




24


these species were represented by very small samples. Therefore, the details of gene frequencies and relative allele mobilities are not presented. The average heterozygosity for Progomphus obscurus was 0.0432 and for P. n. sp. 0.0270. In the non-gomphines, 9 species of libellulids had a mrean R of 0.0491 (Table 12). This is twice as high as the average value in Gomphus. Ladona deplanata, a libellulid, had the highest H (0.1111) of all species surveyed.

The matrix of unbiased genetic distance estimates for Gomphus and Progomphus is presented in Table 13. Table 14 contains a matrix of standard errors for the distances listed in Table 13. There is no formula for direct computation of the standard error of the unbiased distance so the standard error of Nei's standard distance is presented as a close estimate.





25


Table 5. Allele frequencies of Gomphus abbreviaLus, G.
brevis, G. viridifrons, G. apomyius, and 3
populations of a new species of Gomphus.



Enzyme Allele z z z




I I I I W) ~ t 0


LDH N 35 47 5 87 73 12 7 6
100 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
HEX N 35 47 5 87 73 12 7 6
100 - - - - 1.00 1.00 - 1.00
102 1.00 1.00 1.00 1.00 - - 1.00
IDH-1 N 36 44 5 85 73 12 7 6
100 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
IDH-2 N 36 46 5 87 73 12 7 6
100 1.00 1.00 1.00 1.00 1.00 1.00 .929 1.00
121 - - - - - - .071
PGM N 34 46 5 85 72 12 7 6
91 - - - - - 1.00 -
100 1.00 .989 1.00 .994 1.00 - .857 1.00
107 - - - - - - .143
136 - .011 - .006 - - -
ADH N 35 47 5 87 75 12 7 6
89 - - - - - .083 -
100 1.00 .809 1.00 .897 1.00 .875 1.00 1.00
138 - .191 - .103 - .042 -
TO-1 N 33 46 5 84 72 12 7 5
100 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
TO-2 N 31 45 5 81 68 12 7 5
35 1.00 1.00 1.00 1.00 .007 - -
100 - - - - .993 1.00 1.00 1.00
GEN PRO N 36 46 5 87 75 12 7 6
100 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
ME N 32. 43 5 80 71 8 5 4
53 - - - - .028 - -
100 1.00 1.00 1.00 1.00 .972 1.00 1.00 1.00
GPD N 35 47 5 87 75 12 7 6
57 - - - - - .042 -
100 1.00 1.00 1.00 1.00 1.00 .958 1.00 1.00
PGI N 24 29 1 54 60 8 5 4
45 - - - - - .063 -
47 - - - - .008 - -
61 - .017 - .009 - - -
67 - - - - .075 - -
100 1.00 .931 1.00 .963 .917 .937 1.00 1.00
160 - .052 - .028 - - - -














00'T 00 00' 00'J 00'1 001 00' 00'T 00T
S 8 E L 89 T I1E 9 z N U'iYi
- - - - SOO' - 11 - OPTI
- - - P7TO - -
- - 00' - - - - - LOT
00'T OO'T - 986 TL6 00*1 996 986' 00T
- - - - ETC - T10* PTO*
- - - 0 - 0 - 9
9 L MT 1L L8 S 9P7 9E N zis
- - - O- 9- - - - '901

S - - L6' 00 00 -T 00T
- L0 O - 86
L99 L 0 00 - -' -* - S6
- - - 00 - - - - - E6








-0' -0' 00 - 6L6' 00LT 10 T L000
- 0' * - T 6 00' 606 006 06
80' L - - - - -
- - - 900* - - 9 OL
9 L TT CL 88 S 6P' VE' N 1-ILSH
EE'0 00'T 00'1 OU6 001* 00'T 00'T 00'T 001
9 L 61 69 99 9 EP TE. N dv H

9 L ET SL L8 S LI? C N -HdDV

9 L. ET SL L8 9 LI? 9E N 1-HldDV
00'T 00T 001T 001 00'T 001T 001T 0T 001
v? 9 6 9 EL S 8 0 E HCD
- - - 0 01 - T OGLI
001T 001T 001 - 6L6* 00.1 00T SS6' 001
- - - - 1 :0 S P 90 L9
S L ET Iz T 1 11 N 02H
001T OOMT 001T 001T 001T 001T 0 OO*M 001
z E ? v~ H ? 1 6 VT N ){QV
L99 *- - - - - - - 1T
001T 00T 1L6' 001T 001T 0T OUT 001
- - - 60 Z 0 L8
E 9 6 69 6L 9 CI? 1 N HCI-X3IH

:7; H >

zl CIA Cl

Cl))
a4iuvei~u


panUT-4u0D -9 @iqei






27


Table 6. Allele frequencies of Gomphus parvidens, G.
lineatifrons, G. dilatatus, G. furcifer, G.
pallidus, G. villosipes, G. townesi, and G.
potulentus.


Enzyme Allele t
> H H H: 0 0



LDH N 1 10 1 10 24 1 8 10
80 - - - 1.00 1.00 1.00 -
100 1.00 1.00 1.00 - - - 1.00 1.00
HEX N 1 10 1 10 21 1 8 10
97 - - -1.00 - - -
100 - - - - - -1.00
102 1.00 1.00 1.00 - 1.00 1.00 - 1.00
IDH-1 N 1 10 1 10 24 1 8 9
93 - - - - - - 1.00
100 1.00 1.00 1.00 - - - -
109 - - - - - - -1.00
122 - - - 1.00 1.00 - -
130 - - 1.00 -
IDH-2 N 1 10 1 10 24 1 8 9
60 - - - .150 - - -
100 1.00 1.00 1.00 .850 1.00 1.00 1.00 1.00
PGM N 1 10 1 10 24 1 8 10
98 - - - - - - 1.00 .950
93 - - - .100 - 1.00 -
100 1.00 1.00 1.00 - - - -
102 - - - .900 - - -
107 - - - -1.00 - -
124 - - - - - - - .050
ADH N 1 10 1 10 24 1 8 10
81 - .900 - .800 - - -
100 1.00 .050 1.00 - 1.00 - -
105 - - - .200 - -
138 - .050 - - - - 1.00 1.00
163 - - - - - 1.00 -
TO-1 N 1 10 1 10 24 1 8 10
100 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
TO-2 N 1 10 1 10 24 1 6 8
35 1.00 - - - - - -
48 - - - .300 - 1.00 1.00 1.00
100 - 1.00 1.00 .700 1.00 - -
GEN PRO N 1 10 1 10 24 1 8 10
88 - - - - - - 1.00 1.00
100 1.00 1.00 1.00 1.00 1.00 1.00 -
ME N 1 6 1 6 18 1 8 8
50 - - - - .722 - -
80 - - - - - - 1.00
100 1.00 1.00 1.00 - .278 - -
133 - - - - - - - 1.00
222 - - -1.00 - - - -











- 0 - - - - - - 9T

- - - - - - T

- - 0001 - - - - 68
- 0 * T - 98

01- - C-0 - - - - 16
0I- - - 00'0-8' - 06

- - - - - -0' 00 T - 98T
6 8 T zI T 0 1 - 1 N Z: -LS
- - LTL - - -6E1
- - 96T v ET-0 - 01 T - 911
- ' 0 0 0 0 00'T 001
L - - 80 - - - - 96
01 T -H - E6
- OO ' ' ' 00'T 008* - 06
- -- - - os oo900 - V28
01 8 1 HE 01 T 01 T N T-Hd9
- - - 00'T 01' - - 90T
00T 00'T 00 00 - - - 00 001
8 z T~ 01 1T 01 1 N dV1I
- 00'1 00T 001T 00'T 00'T 001* 00'1 001
00 - - - -- - - E6
01 8 1 I OT T 01 T N H-HdDV
00 0 - 00 001 009 - 00'T 00'1 00'T 00T
01 8 - - 01 T 01 1 N T-HdDV
001T 001 0* 00 T 00 0C 001T 00*T 00T 001
9 9 1 TE 01 T 8 1 NH1
- - 001T 00'T - - - - BLT
006- 00 - 00'T - 001
L 9 T H~ 8 0 9 0 N 181-ET - - - - - - - 111
SL8' 00 01 001 001 00-- -1 -01 001
v 7 v T 7 v Z: LI N m>iCv

VVt6' 00'T 001 9L6* 001 001T 00"T - 1T


6 8 1 TE 01 T 8 1 N HG-X211l
- - O00T 798 - - - 0017
001 - - 9171* - 001T OCT 00T 001
- - - 001T S 6
- 001T - T 9
8 9 1 17E L T 9 T N i~d
- - 001T 001 - - - OLT
006* 00*1 - - 00*T O001 001T 001
- - - 0011- 9L
001' - L 9
01 8 1 17E 01 1 01 1 N QdD


u 4


panuTquoD *9 O~qij



































































001T 0027 0027 0027 00*7 00'T 0027 00'T 001
01T L T v z 8 1 9 T N UIT1v

n 4 fZI
EH z I-q@TTTV @uiAzuH
0 0 HD H H
r H E- 1 0A m - 2


p~flu~uoD 9 Gic{pi





30


Table 7. Allele frequencies of Gomphus plagiatus,
G. spicatus, 2 populations of G. exilis, and
2 populations of G. lividus.



Enzyme Allele
x' LH H- C


H I I + I I + C) ) NJ tj H N NJ

LDH N 41 10 16 3 19 2 12 14
80 - 1.00 1.00 1.00 1.00 1.00 1.00 1.00
100 1.00 - - - - - -
HEX N 42 10 15 3 18 2 12 14
100 1.00 - - - - - -
102 - 1.00 1.00 1.00 1.00 1.00 1.00 1.00
IDH-1 N 44 10 15 4 19 2 12 14
100 - 1.00 1.00 1.00 1.00 1.00 1.00 1.00
109 1.00 - - - - - -
IDH-2 N 44 10 15 4 19 2 12 14
73 - - - - - - .042 .036
100 1.00 1.00 1.00 1.00 1.00 1.00 .958 .964
PGM N 43 9 15 5 20 2 14 16
88 1.00 - - - - - -
91 - - .500 .600 .525 .250 - .031
100 - 1.00 .500 .400 .475 .750 1.00 .969
ADH N 44 10 18 3 21 2 14 16
81 - - - .167 .024 .500 - .063
100 - - 1.00 .833 .976 .500 1.00 .937
138 1.00 - - - - - -
163 - 1.00 - - - - -
TO-1 N 41 10 16 3 19 2 12 14
100 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
TO-2 N 39 9 13 3 16 2 10 12
35 .039 - - - - - -
48 .961 1.00 1.00 1.00 1.00 1.00 1.00 1.00
GEN PRO N 44 10 16 4 20 2 14 16
100 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
ME N 35 10 12 3 15 2 8 10
45 - 1.00 - - - - -
90 - - 1.00 1.00 1.00 1.00 1.00 1.00
133 1.00 - - - - - -
GPD N 44 10 15 4 19 2 12 14
72 - 1.00 - - - - -
100 1.00 - 1.00 1.00 1.00 1.00 1.00 1.00
PGI N 32 9 14 4 18 2 12 14
53 - - .036 - .028 - -
61 1.00 - - - - - -
100 - 1.00 .964 1.00 .972 .750 1.00 .964
425 - - - - - .250 - .036























00T 00'T 00'T 00'T 001T 001T 001T 00'T 00T
LIT 91 1 z 9 LIT 6 91v N U'1iv
- - - - - - 00 T TLT

- - - 00T 001 00 - - 00T
00*T 00'T 00'T - - - - - 98
91 VT1 z T: 9 91 01 17 N zi
- - - - - - - 1 8 86
99T' ET 093 - - - - - 96
- - - 090' - L90' - 6LT' 6
L9Z* 69 ' 09* 096* 001T 6 001 - 06
L9S* 8LS* 009* V -18
S1 ET z 0z 9 91 01 z17 N T-1LSR
001T 001T 00'T 00T 00T 00'T - - 90T
- - - - 00 00 T 001
VT1 ZT z 61 v7 S1 6 9z N dVT[
00'1 00'T 00'1 001T 00'T 001T 00'T 00*T 00T
91T 7 z T t LTI 01T 00 N -HdDV
00*T 00*T 00'T 00*T 00*T 00'T 00*T 00*T 00T
91 VT z 1Z v7 LI 01 17 N 1-HdDV
001T 00'T 00'T 00T6 00T 00' 001T 00'T 00T
N1 T1 1 81 E ST 9 17 N HCID
00'T 00'T 00T 106* 00'T H6' 00'T 0016 00T
- - - 690' - LLO' - - 1TL
TT 6 L LT v EE6 8 0z N C1H
- -- - - - - 8ZO 171
001T 001T 001T 001T 001T 001T 001T ZL6* 001
L 9 1 Z17 E 6 L 8T N NOV
- - - 090' - L90* - 00'T 1ZT
OOMT 001T 00T 096* 00T EE6' 001T - 001
91 PT1 z oz 9 S1 01 91 N HQI-XaH

(N4 C14 '-i (4 (\I -i u u + I I + I I H
a4H a4 0
co 0
H, H
H ~1~1w


p~nT~u~J L oiqpl


UE





32


Table 8. Allele frequencies of Gomphus hodgesi, G.
australis, G. brimleyi, G. diminutus, and 2
populations of G. cavillaris.



Enzyme Allele




801.0 a.0 - -1 -v 1 0
+
0 HLDH N 4 1 31 21 10 31 62 9
80 1.00 1.00 - - - - - 1.00
100 - - 1.00 1.00 1.00 1.00 1.00
HEX N 3 1 29 21 10 31 60 9
102 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
IDH-1 N 4 1 32 21 10 31 63 9
100 1.00 1.00 - - - - - 1.00
147 - - 1.00 1.00 1.00 1.00 1.00
IDH-2 N 4 1 32 21 10 31 63 9
100 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
PGM N 4 1 32 21 10 31 63 9
91 1.00 - - - - - -
100 - 1.00 1.00 1.00 1.00 1.00 1.00 1.00
ADH N 4 1 32 21 10 31 63 9
100 1.00 1.00 1.00 .976 1.00 .984 .992 1.00
138 - - - .024 - .016 .008
TO-1 N 4 1 33 21 10 31 64 9
100 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
TO-2 N 4 1 33 21 10 31 64 9
48 1.00 - - - - - -
100 - 1.00 1.00 1.00 1.00 1.00 1.00 1.00
GEN PRO N 4 1 32 21 10 31 63 9
100 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
ME N 0 1 32 21 10 31 63 9
100 - 1.00 1.00 1.00 1.00 1.00 1.00 1.00
GPD N 4 1 32 21 10 31 63 9
100 1.00 1.00 1.00 1.00 .950 .984 .992 1.00
155 - - - - .050 .016 .008
PGI N 4 1 32 21 10 31 63 9
100 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
HEX-DH N 3 1 29 21 10 31 60 9
100 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
ADK N 3 1 23 9 10 19 42 9
100 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
HBD N 4 1 32 21 10 31 63 9
100 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
GDH N 1 1 29 21 10 31 60 9
100 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
ACPH-1 N 4 1 32 21 10 31 63 9
100 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00




































001T 001T 00'1 00*1 00'T 00'T 00'T 00*T 001
6 v 9 ICE 01T T~ EC E I N cvriv
00 - - - - - - VT
- Z66* V86' 0S6* 001T 001T - SE~T

- - ' 9-O 00 - - -G - E61
' - - - - - - O-L 001
- 860 96' O06 - - 0 - E6
- L- E - - - T 18
6 IC 0V OT T CC T FI 9:4

S L8 - - - - - - -86
- C86' L96' 006' 00' 00GT 00'T '6
- L1* CEO* OOT' -oS Z6- 0
- -' - - r 0 98
8 6S OE 01 0z 6Z 1 NT-I2
0O*T O0*GO 00 0 GM OT 0l OOMT 00' 901
S LS IC 01 IZ 9z TI N dv'1
00*1 00'T 0GM 00GM O0* 0GMT 00M OM* 001
6 E9 IC 01 IZ UC I N Z-HdDY


2 H + I I L)
H--1 Q4
+ 4 1
u
u UTV w u


panuqUOD -8 GTqL1


CCE










Allele frequencies of minutus, Progomphus n.


2 populations of Gomphus sp., and P. obscurus.


Enzyme Allele


I I+ FIn


LDH N 28 2 30 4 18
80 1.00 1.00 1.00 -
120 - - - 1.00 1.00
HEX N 17 2 19 4 16
92 - - - 1.00 1.00
102 1.00 1.00 1.00 -
IDH-l N 28 2 30 4 18
86 - - - - 1.00
88 - - - 1.00
100 1.00 1.00 1.00 -
IDH-2 N 28 2 30 4 18
73 .018 - .017 -
100 .982 1.00 .983 -
160 - - - 1.00 1.00
TGM N 28 2 30 4 18
92 - - - - .056
100 1.00 1.00 1.00 -
103 - - - 1.00 .944
ADH N 28 2 30 4 17
68 - - - - .029
75 - - - - .971
100 1.00 1.00 1.00 -
163 - - - 1.00
TO-1 N 28 2 30 4 18
56 - - - 1.00 .941
100 1.00 1.00 1.00 -
167 - - - - .059
TO-2 N 28 2 30 4 18
34 - - - 1.00 1.00
48 1.00 1.00 1.00 -
GEN PRO N 28 2 30 4 18
88 - - - 1.00 1.00
100 1.00 1.00 1.00 -
ME N 20 2 22 4 15
42 - - - .125
100 1.00 1.00 1.00 -
155 - - - .875 1.00
GPD N 28 2 30 4 17
100 1.00 1.00 1.00 -
170 - - - 1.00 1.00


Table 9.


34










Table 9.


Continued.


Enzyme Allele







56 - - - 1.00 1.00
61 .179 - .167 -
100 .821 1.00 .833 -
HEX-DH N 27 2 29 2 13
95 - - - 1.00 1.00
100 1.00 1.00 1.00
ADK N 28 2 30 2 11
100 1.00 1.00 1.00 1.00 1.00
HBD N 27 2 29 4 17
78 - - - 1.00 1.00
100 1.00 1.00 1.00 -
GDH N 16 2 13 2 14
100 1.00 1.00 1.00 -
108 - - - 1.00 1.00
ACPH-1 N 23 2 30 3 12
100 1.00 1.00 1.00 1.00 1.00
ACP-2 N 28 2 30 2 12
100 1.00 1.00 1.00 1.00 1.00
LAP N 28 0 28 4 11
100 - - - 1.00 1.00
106 1.00 - 1.00 -
EST-1 N 24 2 26 4 18
90 .333 .500 .346 -
95 .667 .500 .654 -
117 - - - .750 .111
127 - - - .250 .194
137 - - - - .472
155 - - - - .223
EST-2 N 28 2 30 4 18
107 1.00 1.00 1.00 -
260 - - - - 1.00
460 - - - 1.00
KLD N 28 2 30 4 12
100 1.00 1.00 1.00 -
126 - - - 1.00 1.00


35

































100 --


origin__o n
0 0 F o 0 o U)-


0- CD o H --> 0 H 0 9 < H0 H- rt 0 F-- H 0 0 H CD U) ft 0 A'<
0 Fj H
) H
U)


ft
0 H> U)


o P.


(D D, m r
- Un


CD Un T'0 lc F-- a Qt Fo Il F-D 0


U) ct
U) t (o c
co


SI
D

(D F-


0 Ft o M- HCD
U)


H--D ' o
F- 0fI 0 0i Cl <~ 9

ft H-o D 0

U) j U)
0
co


<
H
H.

0
H

In F
0.
0
U)


CD
U


(D
(
( 0 U)-


Figure 1 Ra a Live a 1alele mohi 1.1 t i a s af ,c id phosphatase-1
in species of Gomphus and Proqomphus .






37


95--
100 -"


origin


o o- o>0 :-r
0 0 0 pP3 P - 0< I-. m o


H- (m


o to 'o 'U N In (I 0 H- P- C> rt

H- 0> H- H' to rt 0) (D
0 ct 0 co crt
on r
to


U H- 0o 0 H- H. Ve F-' F -' C) m- T
(D 0 H- M P - HU) - tl to to to ( - 0- rt (> H H- ( () rt 'U t-co '1 0 tH
0 to aHa t
to 0 o
So
"Cl
(D


Figure 2 Rela t i "'e a llele mobilit i es r f acid phospha ta se-2
in species of Gomphus and Prougomiphus.


od e


to C o So


H- H
- H

tnrt 0 to ( Un


m 'c
0

to


-1 tr ~ H- Hi

-
Qa H H- t

0
to
:5
in


to tr

t o












































100-- ------------111~~



143









origin- - P-s > . I'l "C 0
m - U) 2 < P- H- N rt o 0 Z H-H- P rta H- I-, Cn tir - J 1- - I (D Ca P, - W> P- F LOr) (CH~D WU) rU)t OJ D
C U) U) W FDJ K H : - H- 0 (t 0
Fj- U)H- U n r
U)o U)


Figure 3. Relat-ive allel- mehilit ics of a(]nylate kinase
in species of Gomphus and Pro';omphus.


33


(
0
CD U)


- C Fi
H- D> 0

0 HH H - 0 C

'0U) H,
(D
-D
U)


0


ri r0

[-'0 Di H- I i '0


Ct Q, HH- (D C

ho U
0
0I
rf)



P. F- LT, H- e I CD CY


0 (t
(n0



















68 751." 81-

89

100-
105-


130-
138



163








origin


0 U' W - -' (D C(n '?0 '0 i < 'C t t2) 0 -'C5 Wi :lC 5 P_ Fl- m Fl > r 0 o P_ X rI d'o 0 0a ' P> w ro
Q - P-- P- DC - rt0Fl - o r H-Det :-i ( 0 C) C.l H C--Q r :D CO)s) 0 r t
!D- -t ( M D Ha -p ) -e ( 0 - rt W Ha - O -(D TI
m 1: (D r) (> ) U)- - t l (D s- rn[1 D r - (). P o H 0 (n U (- M - - n r- rt '-I () rt Fl (hD r- "i
n P)- n U l) rn 0 ( 0 (t
C E U o o 0 n C












Figure 4. Relative allele mobilities of alcohol icehydrogenase in species of Gomphus and Progomphus.


111)


































100-


126-


0

oW
So Ce
C


- H pi H C 0 H- D M
Cl j Fl- rtF0 P- 10C

Cl C D CD CD C( C rt


U P- UD
UM


Figure 5 Re li aLive all ole mobil i t es of a idolase in species
of Gomphus and Progomnphus.


Fj )


origin


H-Cl
0) (l



C rt


rt < ht,


(D 0 P- P-n n CD h+i
(D
'C (D ) (A


rt


C


(D <1 CD P-' r-t

rt 11 C-D


0
In


0 U)
0
CD


(D UY

(D























70
94 36 90 92 93
95
98 100


106115 117
124 127 129 137

155


origin


0 1-0 0,n C) -o :rso) F ( o T e t < T I-t) t p> o - - P.O H- H-0 F- 0 0 1- Q C - 1- N
0 V, P- -'--I rte ,Q -s 0 Q r, F o ( > D (D ur- (tD H P.a W H.-">0- O --H rt 0)H.K
F- HOU )U)r U)CC C rt & (D W (n , h Wa ct a H
(J) rt L> - F -H.(n f t j P- - (D cf - H C)r 0 C5 r V H - U (n t 'o j) " f lfh II (f) ' M - H. ( U) G (D ( FI mu
r (n In U) 0
U)


Figure 6. Relative allele mobi.ities of esterase-1 in
species of Gomiphus and PrgIomphus.


41


"I - ( 0 (
H C


....-_=-

















54

70 81
8 6 T 89
91 93
100-
107-

120--- = 124- 125 13 5

140 144

171



260



460 origin


0
-'


C)

U
Hj
o U)


U)


n n U -,v Fs- D un i
r, ) ~~ 0 x T F-' 0 P < - P ) rt _Q H-H-Ut) rt
Zt F-j (Ds Wj in r. H- U) w (D S-t D ) U) P. U) f t U) C r U P. P. m r t U+ > U - (n
U) U)--o t


o H-t ) c (D 0 P-HF(D
U)


U, H-"0 U)F1 (D U '

r+ 0- . i, -'e U) (t H-P H- ) ft --D U F HU) I U) 0 ft
aI


Figure 7. Relative allele mobilities of esterase-2 in
species of Gomphus and Progomphus.


42


--






43


88



100 -


origin_






-
U)e


n o (T P 7 71 C >)- u r w r-c-r
F- s-U H c 0 H-x e e-0 H 0 - t)> ~c H- H- Ce e0S5 - H-U) Q o< H- H- > rt P ( H H - 0 H 0 P

r 0 - D> ) o rt D> (D M U) o h > rt (I P- - H Srt > U)H P ) U)n (nr D U)HFIU 0
In U) U) 0 0
0 U)
U)o


Figure 8 Relative allele mobilities of general protein
in species of Gomphus and Progonphus.


H O'

P- (De
H
rt
0
U)




44


100 A
108-,_


Ha n y 0> H- < n-m
F-- U)

cH FD 01 rt 0J- F-'H hn - HU)- U


03 (D 'U F'U0 0 s.N 1 1 0
0' < H- H- 0> cief H- H' H a H CD H- > H- P (O r In rt R) (D


L n fH-U) -tco


(D 0 HH- H- H oH U (D
In


H D- H- (.' SF- 0 Fl
() 01 C < I - rt- 01 HCD ft H- C[D H H H' 01

[0


'0

H
H U)


F

P. in
Fl C 0

H-U)n bh
0 01
U)


0>01 '*

CD 'U
H
0> ct
H U)


Figure 9 Relati ve a I1 e mobi L i ties of c lu-amate ehydrogenase in species of Gomphus and Progomphus.


origin


0 '1
-: - (t o 01 H

in r o U) w



















57 67 72 76 100




148 170


origin

0 n -

U)
C)U)


z 'c
U)


-1 - a eO~C a -- C U) 'L 'C 4 'C> c -a-'C1 C
l 3I -1 H-U ca o- -H > rt < F-l 0k - D Ct 5 1 H*-rC*-CCl-HCCLf C --' C, D> U) C U) Ft Di ( [) U) a H D> t > H(n ct P -n r--H- ' P- P-C (D C t 1-- (1) rC H- . (n r) (t (n Fle C t) U)
U) - U) M r (D U) U)
U) U) Un 0
D
Un


Figure 1 0. P oIa Lve a] I-e mobi lit i (s o f a--()I yccm phosphate
(Iehyd]rgenase in species of Gomphus and Progoaphus.


45


U)


a o

-a
F'- U) hi
0 Dl
U)


Di
a
a
H
CD
A H
D> rt
C
U)






















8 3
87

95100-


121


144, -


origin





U) 0
0U)U)
0r- e 0o c U) o


H- 0 ) r 0 P- X IzJ < H- U) a < P- H-H. --E rt0a - s r ri '< P - 0 Hrto< F~-U 0DP-Q

0 a H- H- U)
r- - W
S


- -~

0
mM ( c t 1


0 rt U)
In


W-e T f ii-C


H H C) w 0 0 H- pH- ft

- - 0 rt

(D
m)


Figure 11.


Relative allele moi ] ities of he-anel dehydrogenase in species of Comphus and Pro(omphus.


H - p0 O 0H~
0 e-< E 0 -j

H- ( D H, D U) t- U)
0


H- H 0-H- < hj U) C, H- 0 'U H- ?) <0 (tJ


S0> ro Ut






47


92 97 100


102


or ig in


0 0,
-


Figure 12.


U) UO


H-H.

0 0U rt
0 U)


3' la) iD (D F 0
I r- 0 P. X U>

- rtu -Q o F-n 1- rD W> UI I n 0t H - U-)) U- )


-U U H-' 0

-
to C F t 1 0 c-


rt ' '0 0 P- C

CD 0 PU) U> 0r H-F--C
'0 U)
(D m)


0P- - H- P
C) CD C <
ih C> t Q, (D rt - D
0 0 HU
U) H Un
0 0o


CD '0
0


H
0


< ( D 0


P- HD 'T0H- U) < h H7 rw C


Re a tL i ve a LI e I emobi lit is rf h rxok inase in species of Gomnphus and Progomphus.


























65 74 78


100









178












1700


origin


o u CD Ha:y Fvru0 U) CDCD
D Un U) ft CD-C CD
CD m
U)


DJF D0 X- IT! "dh 0 <~ H- U C, < P- H-a) >t H-S rt QF---CD PD WD r' - n U) (-t W (D

inH-H- toCDr+
H- U) U) C
U) U)


Figure 13. Reltive allele mobiIities of b--h"-]r:-:utyrate
acid d(eby'lrogenase in species of Gomphus anCd
Progomphus .


CD U)


0H F- M -t U) 0 > H- C D rt(

(


H D o
o a



o>M rt OHs
- CD CD

aU)


H
0,
(

a
0


(D 0' < a U) H (D 'H
Di
rD
U)




























86 88


9 3 100 109



122


130


147


origin -77

P- ) Q < I- H- OC) Ft
o -5 r- H- - rC) -- F- 0 O r, U r t -i (D n . P - (n r- tU) r - ( -( 0 [) * rt UJ -t
n P- m In C e
(n o
c)


Figure 1 4. Rl at i v(, a I I ele moibi Ii Lies of
dehydrooenase-1 in species of
Progormphus.


i -l), i fra dt Gomlphus 'Ind


49


o0 H


'C U
o


0


(n
F-
Li U)


'C)


(Drt P> (D r-> 1t1 o, (D rt H- 0
in f, m
0
(n


-- T
tl (D L,' 0e - 0



o F
0 Li
U-) U)




















60 73 100-


121 160


















origin


0

3 F-


Figure 15.


Q Dao 0H -H-0 Ip -1e H0 -H- H rte a -a
0 0 IsrD Un rt 3' L
hj H*U)P
Un


S03' F- (D UI) '0 '0 r 0 H- < X -' 0 m 0a -s 0J U) 0 ( rtD CD o H -*) P rt H- Ort
U)


(I
0
CD
U)
H-


-VC 1 0)


H 0 Ct ct P- rJ ( Ut U)
(-D D t (n(n


I In 0 <
C) 0-e H. [' CD 0n Fo U
0
*1


0< H H- U)
F-4 0aH 0-(n U)


0 0 C)-.



H

rt
0
()


Re I at i -e ae 1 I e mobile Li s of i soci ra e clehydrogetase-2 in species of Gomphus and Progomphus.


50




























80 100


120


origin_- O
-u

Sa
U--


Figure 16.


S
H
F)
a
rt
a
en


C -I D N
-- l e

a i-s D
rt CD '< H a i H- Hrto
(n- In


0 P- x Co < H'0 H H

-mw


CD H f-t C a ft
(n


171
0 (t

a (D

a
'I


ft
0


en


0- a -- 0 - 0 -0 Hs o a CD <5 o s H->H ot - D et P- (D l en HM a r h ) U) o o
en) 0
F) (n


H- (D


-m -:
-h H'
o <> m e


Un
Uo


Rel active all ele mobilii 0 nf Iact-te dehydrogenase in species of G-omplhiUs 7nc Progomiphus


--






52


42
45 50
53




80

9o----100


133 155



222


origin____(7





C H U)


U


C ct CU
U)


CU, H-.
CU H
CU
C rt
C U)


n Op w pU CD U rt c' ( u U Wd H C 0 H- r 0 X 0 H- 0
:-- rt - H 0 ) rfl CU I-- " ( , P H- C- H- F-- (D 0 PD W ) n r ) rt WU CD M c oC 1-1. H --. C rt- H- - C F- P- U) C rt- 'U (n
p- U) (n (D
W U) m


Figure 17. Re]ative allele mobi l iti rs of malic cI)znyme in
species of Gomphus and Progomphus-


I h CU
H
a H
I- s
C)
H


I- H'


rt CU
U)


tT (D
e


F-' 0 0U H- CU 'r


M MO rt C, HH- CD C Vt, CU U)
H U
0
(n


F-
H

F
Hi
H
0 CU


C> U r3, HU'
F A
H
CU rU
C























88 91 92 93 100 102 103 107



124 136


origin

- n - Ho - Z H-u (n ftu r O'- rt

(
(U


Figure 18.


n 0 > :(D L L O H P-j 0
H'(U H Us

F- H- HF- (fl U)


H-' (D 10 'U 'TI H- ::'0 H' 0 F- H- a it
Qa H- H> H- F-M (n rt P) (D Un O (-


00 N
o H- PJ G E, F F-- H (T (D H. ( (1 O H-F(n (n P- H) PH H- (D(

(U
r-js o


F-- H- HJ '0 S(D -1 rn i H- Hi rt P- (U USH, U
0
rn


< trH 0
H ( U' P- -: '- Un 0l H- (U *TF-- U) <: Lb HH H) 0 rt
U M
(I Us


Relative allele mobilitir s of phosphnq lucomutase in species of Gomphus and Progomphus.




















45
47 53 56 61 67


95 100





160











400 425


origin


'0 F-h ' 'CI H C F--- H



(D t-- (D
U) H-rt it H.

) 1 U5


U)


Figure 19. Relative allele nmobilities of phosphf-rlucOsC?
isomerase in species of Gomphus and Progomphus.


54


U) C C)U)U)

C 'Ct hi.
C
(n


0 H-X
()< H- F-'
CD HU)C (n
H- U)


Dn 0 'p 0 H-m 0 F- F 31 < H-o U H- F- ri
F-- F- hjM C F- (D F-i (+-F 0 C H H- F(flH- U)
(U


Ft
0

(D
(U) F--


U) ao ' - 0

(o a C

t (D F t F j r) ( ft
(nC
(0


H

-
0

'F(D
(0


'0
0


C rn


H D (Y
(H-D '(1

I-h H. 0 H>
0 rt o C U) U)





















56


100--


167


origin


f-i
C: tr


Figure .)0.


-1

0)
-)
71


-a n rt 1

0
f-f 0


o T W '>-' (D M) 7 'U w 0 0 H- x " 0
P. U) a < H- H- W
H- ri5 f- P- H 0 Q 0
--H Fj (D (I. P- 0) P- H' H'( 0 U)w 0) w rt N m J P- (n r A
M r-- I t
U) In r


0 -t - 0 0)1 (D 0 H(T)
-U


Cl I-,
H' 01
f-t
0)
rU


0) 'U) 0 h S1 In

U ) Hhi) 0

o 0)
0) U)


CD C)'

-

ft
0r In


Relative al1eI nbi I i t i r t tra at- m oxidasc-1 in species of Gomnplus and Progomphus (cathodlal migrant).


- - - - - - ---------






















3 5


48 -


100


origin__U)
o 0r tU


'tY~P. (n
a -1 H- U )
* rt -l
- r (D N M) Ln (t ( > f la
' 1 H- p
- p H- U)
In


o H- X 'd - 0 0 H- H 0 Os K H- H- 0 rt- 1- I-'
(D H, - H' C 0 o 0 ) 0 Um rt 0 I 0 U U) 0
H- m Ort PH- rU) e gt- 'tJ rn
U) 0 0
U) rn


Figure 2 1 Re 1 t i ee 7i 1 I Cl mobiiies oF- i t ra z'1 i umn
oxidase-2 in species of G;omphus and Progomphus.


~)


a

H
hi
0 1-1


0 lC 0 1- 0. 0o 0


W rt 0, Prt H.O 0
0 h o U) 10 Fl~ In
0 0j
Ln


H- hi 0y Fl (D U-~

, - oD 'T F- rn <

0~
o et
0 0





































100- 106-


or i g in Figure 22 .


o 0



03 0 Sci U)to


H

Ct
0 U)


H
0 i-f
0t
U)


0 J :3'o a rt w> hl r 0 H- X vT 0 0 H- Cf Fl- H- 0 H- h "
< Hu) < HH-P t :- WF-I- I-- 10 11 D Ct H - :-tLO H- 0- : F--C) ci (D < l H-< F U) F, F- VI (D ID H- H- " (D 0 H H- t W> P-k lH-(T C (- D CD U 0 ) wt (n F; n t w (tD uI (A C V- H u) < J~ N- m _ U r, (-D- Cr P-o (D r- Ft .
P.- H- 0 C rf O' U) CrH) M H
U- oO r t- 'TU10-O o- ri
U) wO (D ulflW 0 it
U)U) U) 0 0j 0
En w) U


I e] a t i ve a I Iee Imobi I i t es of leucin n aminopept idase in species of Gomphus and Progomphus.















Table 10. Percent loci polymorphic and percent loci variable
for 22 alleles surveyed in Gomphus and Progomphus.


% Polymorphic

1 2 3
5% 1% % Variable

n. sp. 76 9.1 22.7 27.3
abbreviatus 63 4.5 18.2 27.3
brevis 10 9.1 13.6 13.6
viridifrons 6 13.6 13.6 13.6
apomyius 5 9.1 9.1 9.1
parvidens 1 0.0 0.0 0.0
lineatifrons 9 9.1 9.1 9.1
dilatatus 1 4.5 4.5 4.5
furcifer 9 13.6 13.6 13.6
pallidus 23 13.6 18.2 18.2
villosipes 1 0.0 0.0 0.0
townesi 7 0.0 0.0 0.0
potulentus 9 18.2 18.2 18.2
plagiatus 39 4.5 13.6 13.6
spicatus 9 0.0 0.0 0.0
exilis 19 18.2 27.3 27.3
lividus 14 9.1 22.7 22.7
hodgesi 4 9.1 9.1 9.1
australis 1 0.0 0.0 0.0
brimleyi 31 0.0 0.0 0.0
cavillaris 30 0.0 13.6 13.6
diminutus 9 4.5 4.5 4.5
minutus 28 9.1 13.6 13.6
P. n. sp. 4 9.1 9.1 9.1
P. obscurus 16 13.6 18.2 18.2


1
2
3


Most common allele at a frequency of 95% or less. Most common allele at a frequency of 99% or less. Percent of loci variable regardless of frequencies.

























Table 11. Locus by locus and average heterozygosities of Gomphus and Progomphus.














LON MEX I ON-1 ID-2 PGI ADH TO-1 TO-2 GEN PRO ME GPD


N. sP. P-1 0000 N. SP. P-2 .0000 N. SP. P-1 .0000 . OP. Pi-) 3000

Assn 3000 REV 0000 vIN) 0000 APOM .0000

PARV .0000 LINE .2000

GILA .0000 rURC .0000 PALL 0000 VILL .0000

)0NN .0000 PoTU 0000 PLAG .0000 SPIC .0000

EX P-1 .0000 EX P-2 .0000

EX P1.' .0000 LIV P-1 .0000 LIV P-2 0000

L:' PL-2 .3000

0D00 .3000 AUST .0000 3NIN .3000

CAV P-i .0000 CAV P- 3000

CAV P1.2 .0000 CAV . BSPI .0000

Dr( 0000

1M4 P-1 0000 NIN P-2 .0000

NIN P1+2 .0000 P N. SP. .0000 P. ONSC .0000


.0000 .0000 .0000 0000 .0000 .0000 .0000 0000 0000 .0000 .0000 .0000 .0000 .0000 .0000 0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 0000 .0000 .0000 .0000 .000 .0000 .0000 .0000 .0000 .0000 0000


.0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 0000 .0000 .0000 .0000 .0000 .0000 0000 .0000 .0000 0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000

.0000


.0000 .0000 .0000 .0000 .0000 .0000 .1326

.0000 .0000 .0000 .0000

2000

0000 .0000 .0000 .0000 0000 .0000 .0000 .0000

-0000 .0000

.0799 .0689 .0000 .0000

.000 .0000 .0000 .0000 .0000

0000

.0352 .0000 .0328

.0000 .0000


.0000 .0216 .0000

.011' .0000 .0000

.2450 .0000 .0000 .0000 .0000 .1800 .0000 0000 .0000 .0950

.0000 .0000 .0000

.4800 .4988 .3750 .0000

.0604 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .1050


.0000 .3097 .0000

.1954 .0000 .2257 .0000 .0000 .0000 .150

.0000 .3200 .0000 .000 .0000 .0000 .0000 .0000 .5000 .2778

.0465 .5000 .0000

.1172 .000 .0000 .0000

.0465 .0000

.0317 .0157 .0000 .0000 .0000 .0000 .0000 .0571


.0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .3000 .0000 .0000 .0000 .0000 .0000 .0000

.0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 0000 .0000

.1107


.0000 .0000 .0000

.0147 .0000 .0000 .0000 .0000 .0000 .0000

.4200 .0000 .0000 .0000 .0000 .0000

.0740 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0001) .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000


.0000 .0000 .0000

.0000 .0000 .0000 .0000 .0000 .0000

.0000 .0000 .0000 .0000 .0000 .0000 .0000 .0548 .0799 0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000

.0000 .0000 .3000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .4013 .0000 .0000 .0000 .0000 .0000 .0000 .1800

.0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000

.0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .3000 .0000 ----- .0000

.0000 .0000 .0000 .0000 .0000 .0000 .0000 0000 .0000 .0000 .J000 0950 .0000 .0000 .0317 .0000 .0000 .0157 .0000 .3000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000

.0000 .2148 .0000 .0000 .0000 .0000


W .0000 .0000 .0000 3162 .0484 0781 .0036 .0164 .0000 .0218 .0114 0448 .0251


.0000 .1303 .0000

.0710

.1540 .1172 .0000 .0000 .0000 .0000 .0000 .0000 .2491 .0000 .0000 .0000 .0000 .0000 .0689

.0000

.0541

.3750 .0000 .0649

.0000

.2000 0000 .0000 .3000 .0000 .0000 .0000

.2934 .0000 . 2778 .0000 .3000


.0000 .0000 .0000 .0000 .0563 .0000 .0000 .4444 .0000 .0000 .0000 .0000

.0465 .0000 .0000 .1050

.0000 .0000

.1245 .0000

.0950 .0000 0000 .0000 .0000 .0000 .3000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000


.0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .218

.0541

.0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .3000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000


.0469 .0000 .0000

.0407 .0000 .0000 .0000 .0000



.0000



.0000 .0000

.0000 .0000 .0000 .0000 .0000








.0000 .0000 .0000 .0000 .0000 0000 .0000 0000 .0000 .0000 .0000 .0000 .0000 .0000

0000
0000


.0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 ,0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000


.0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000

.0000 .0000 .0000 .0000 .0000 .0000 .3000 .0000 ,0000 .0000 .0000 .0000 .0000 10000 .0000 0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000


.0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 0000 .0000 .0000 .0000 .0000 .0000 0000 0000 .0000 .0000 .0000 .0000 .0000

.0000


.0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000

0000

.0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000

.0000


.1629 .0790

.4200

.1591 .0536

.0469

.2552

.4861 .0000

.1200 .5000 .0000 .4 390 .0000 .0000 .0000 .2934 .0000

.1245 .0000 .0950 .6250 .5711 .5799 .5000 .0000 .0000 .0000

.1400 .0644



.2144 .4444 .5000

4S27 1750 .A775


.0274 .0844 .0000 .0561

.0266 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000

.0000 .0000 .0700 .0000

000 .0000 .0950 .0317 .0155 .0000 .0000 .0000 .0000 0000 0000


.0003 0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 0000 .0000 .0000

.0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000

.0000


.0126

.0244

.0191 .0239 .0164

.0212

.02IR .0421 .0000

.0230

.0214

.0504

.0017 .0000 .0000

.0272

.0192

.0000

.0416 .0345

.0409 .0402 .0296

.0407

.0417 .0000 .0000 .0021 .3160

.0074 .0016

.0099 .015L

.02)i .0347

.0210 .04)2


PCI MEX-DM ADO


MSD COH ACPH-1 ACPH-2 LAP EST-1 EST-2 AL.)


.0000 .0074 .0000 .0000 0000 .0000 .2159 .0196 .0000






61


Table 12. Average heterozygosities, sample sizes, number of
loci surveyed and number of variable loci in
selected species of Odonata other than Gomphus
and Progomphus.


Species N


Epitheca cynosura E. stella E. costalis E. sepia

Libellula incesta L. flavida L. semifasciata L. auripennis L. needhami

0. ferruginea Ladona deplanata Plathemis lydia Pachydiplax longipennis Didymops floridensis P-1 D. floridensis P-2 D. transversa Gomphaeschna antilope G. furcillata P-1 G. furcillata P-2 Epiaeschna heros Coryphaeschna ingens Basiaeschna janata Anax junius


No. Loci 21 21

21 17

17 16 17

18 18 17 17 19 19

22 22 22 18

20 22

21 13

20

19


No.
Variable

3

2 3 0 1 1

2 2

1 0 5

4 4 2 1

2

0 1 1 1 0 0 0


.0655 .0623 .0397 .0000

.0294 .0312 .0589

.0486 .0278 .0000 .1111 .0848

.0504 .0202 .0202

.0127 .0000 .0139 .0075 .0233

.0000 .0000 .0000
































Table 13. Matrix of unbiased genetic distance estimates
for Gomphus and Progomphus based on analysis
of 22 loci.






























C*D T0 Z




















CD D
* * * * *








o O* NJ ~ J C


































*D
- - 1








*


ON ON


r. .4l N N N N NJ NJ --' -- 0 . O 1 N 4S. . .J . . .N ON . Z .t. ..

C CD co co I t, M OC j Wm





-O 00
- z--. >- o








ON 1O O N OCN ON ON 141 S.
. . O S. . 0 0 . . . 4 N 1 .















m- o--01-m'S.14lN l- ZN4100.--oSa,
14 Lu NJS1 N0 0ON0O ON-4 J uS ON .
10 01 C, 1MN3 WC3m
ON ON 4 1 0 - 4 Om C N CD S 1 ZN O O ON O . . . . . . . . . . . . . . 0 . ZN




w Uu m 00 w I0 m wONO4NNJOONON CO ON S. O C O J -J 0 S O O I ON 0 N . . .. - . . .-0. . . . .





14ON 14 ON ON O O N 4 o -4 - O 4 O 10 w CO ON 0- .1 C0 - ON












Ln Cw U - - - - - t W M
CCDN-ONO OZON- - , a, m O O S - O 0 S CO ON J . . ON SJ .O -4
NJ 0 ON ON.1 O NJ 14 00-4W ZJN


-4 O ON 141 U O ON ON ON -4 ON O- ON

ONO 0 ON ON O 141 ON 0 4 ON ON O O





1 0 ON O O S 14 ON l 0 S
-4 ON tO NJ NJ 0 1 .S o NJ NJ NJ NJ I...


NJ ON l-'lj ON 0 NJ 01 ON 0 03 O
ON -4 -J ONO CO 0 N J 4 J ON -. Z


S.141 NJ O O O 002-J O ea
S. 1411 0i NJ O O ONU 0000WH



NJ ONONI-J1-NJ OA A W JOOOMNJM




CD s. 1 01 o C ; J M

N .. NJ Lu O O
0 -. J-0 NJ ON 4 . NO
ON 0 CN N 1 O O O

* OON- ----C


C O Cl-N000ON00


ON -4 4 ON NJ 1 -'ON -' ON NJ 4 NJ - N - ON 141 O


-'0ONQ ON -J ON 141 S. NJr 1 0 O ON 4 ZN .1. NJr 10 O O J ON 141 S. L N







- Z - .- -> -.- - - - - - ' - 0
< x x
> m

,.) "I I I I I - 11 I I I I - > . C 0 0 :_ 4 0 0 1 0 I I I






. . . . . . . . . ON . 0 - . - C O - O * - - - - - - '
c, MN w j Lnmoo eoo m a, w o CRfo
41 C S. N L u 00 -411 l JL O N NJ f 5 ON O 4 ON ON ON S. O Lu NJ 4 O

O O - NJ CJ NJ NJ NJ NJ NJ 0 J NJ 0 141414 ON ON ON - NJ 0 01 NJ N 0
11 NJ S. ON O Os 141- NJ ON 14 - '00 N ONI0 \ J -4 ON A 0 W ON o






-4-. -4 . . . S . . . . . . . . . . . . . O O . 0 . . . ON
--..---- -4-4- 4--N 4 S. 141NJ OON 14J NJ 4 S. ON J 4 ON ON ON Or




. J O O 0 S. \ 1- u C 0 0 141 S ON 1J S 4- O ON - O O O S - D



ON ON 0 .0. . 0, 0. . . . . . . . . . . . . . . - - . NJ
s4 ON Nj W I- - - - - oONUi N J ON ONO O NJ O O O 00 SJ 0 . 1-C -4 S. 0 ON ON O O S. 14 1' ON 0 ON S. 0 J -4 ON NJ 00 \J 0
ON ON 10 S. 0- 0 -4 S. ON 0141 S.C ON 0 UJ O C0 S. ON NJ O


ONO .4 NJ . .J NJ .J .J 0 S. S. S. S. 14 14 ON .4 N ON .4 J .J -

C J ONONON- MON100C01'O NO NO NNJ u DON1 O N S.

14 NJ 14 ON O O ON O O O ON - ON O O S. OAe 414 0







O ON - NJ - - -- N N 0 N 0 ONO N 141 .4 1 NJ 0 0
ON -4 ON . . . 14 . . 14 ON 0 -4 ON .N ON 0 ON ON . ON .J ON .
ON 0 NJ O ON- ONa 0 O.N ON - NJ N J NO - S. 000C O ON A ONa - 40
NJ S. - NJ 0 \O N 0-' J - O -40S ONo O ON NJ NJ 4 ON

ZNZNN L u NJN NJ NJOSW4- . -'00 S. S. O O 14 4 J - O . ON 1 0 O .0. 0 0 . . 0 . . . . . . O . . . 0 O



ON A O N -4 S . l 'NJU1NJ... .041 s NJO-.NJO ONO



Z0 CD m S 00 NJS ON ON 'o N m41 O 1 Lu
.u . - ZN ON ON . .ON . . 0 ON 00141 . .ON . ON .
S. NJ -4 41 4 ON N.J.J-J jO NJ .0 ON 14 1 .. ON 0 ON 0J
U-1 NJ a 4J 0 O ON 41 N - 0 4J 1 S. ON 'OO 0 O N



NON . NJM N NJ NJ r NJ 1. 0 J NJ o 0 S. 41 14 ON ON ON 041 ON -4 O 00W S. S. S. - O N NJ 041l O O 0 ON -J O ON 41- C 4J ON I14 N J O O O S. - S 1- O 0 O 0 0 O
N 141 0 '1 O5 NJ 4J 0 G) O O - - N 0141 0 -J -4 O



ONON,..-- - - - MMNJ NJNNJJS0.ONO N mONONaS.4 1 0040ON O OJ ON ON O - ONM S. NJ .N N0 O NJ-J
X41 wNOO-4NONO0 0104S.S.O 1ON NJ00b


ON


11


U1


-1


i-
sJ

































Table 14. Matrix of standard errors for genetic distance
estimates of Gomphus and Progomphus.





65


1 2 3 4 5 6 7 8 9

1) N. SP. P-1
2) N. SP. P-2 .0017
3) N. SP. P-3 .0183 .0213
4) ABBR .0998 .1032 .0986
5) BREV .1173 .1183 .1153 .1026
6) VIRI .0635 .0654 .0599 .0842 .0855
7) APOM .0845 .0863 .0821 .0662 .0757 .0546
-8) LINE .1167 .1137 .1208 .1436 .1429 .1111 .1100
9) FURC .2557 .2564 .2398 .2402 .2395 .2404 .2210 .2231
10) PALL .1981 .1999 .1978 .1987 .1968 .1722 .1804 .1957 .1578
11) VILL .2169 .2151 .2163 .2380 .2157 .2165 .2192 .2158 .1841
12) TOWN .2002 .1903 .2075 .2164 .1959 .2165 .1815 .1982 .2491
13) POTU .1969 .1910 .1844 .2405 .2173 .1981 .2016 .1967 .2498
14) PLAG .1952 .1862 .1910 .1956 .1762 .1939 .1626 .1934 .2192
15) SPIC .1342 .1312 .1396 .1816 .1798 .1486 .1683 .1651 .2063
16) EX P-1 .1072 .1081 .1109 .1512 .1537 .1197 .1379 .1329 .2389
17) EX P-2 .1080 .1089 .1129 .1547 .1536 .1215 .1408 .1310 .2411
18) LIV P-1 .1295 .1300 .1313 .1685 .1534 .1301 .1515 .1117 .2296 19) LIV P-2 .1241 .1248 .1260 .1618 .1596 .1261 .1464 .1196 .2473 20) AUST .1169 .1182 .1148 .1336 .1198 .0880 .1091 .1124 .2417
21) BRIM .1169 .1182 .1148 .1336 .1198 .0880 .1091 .1120 .2415
22) CAV P-1 .1170 .1183 .1149 .1338 .1199 .0881 .1092 .1103 .2367 23) CAV P-2 .1145 .1159 .1112 .1318 .1197 .0874 .1088 .1114 .2398 24) DIMI .1151 .1163 .1132 .1322 .1326 .0976 .1178 .1107 .2407
25) MIN P-1 .1083 .1088 .1107 .1475 .1224 .1067 .1277 .1214 .2462 26) P. N. SP. .4610 .4602 .4609 .4608 .4604 .4602 .4598 .5472 .4588
27) P. OBSC .4603 .4595 .4303 .4600 .4597 .4596 .4592 .5467 .4581




10 11 12 13 14 15 16 17 18

11) VILL .1377
12) TOWN .2582 .2390
13) POTU .2527 .2397 .1512
14) PLAG .2335 .2175 .1161 .1291
15) SPIC .2076 .1816 .2182 .2175 .2175
16) EX P-1 .2006 .2110 .2138 .2121 .2114 .1225
17) EX P-2 .2068 .2147 .2147 .2159 .2139 .1199 .0014
18) LIV P-i .2078 .2090 .2226 .2146 .2087 .1223 .0571 .0568 19) LIV P-2 .2056 .2161 .2286 .2172 .2158 .1242 .0555 .0577 .0125 20) AUST .1965 .2182 .2390 .2194 .2167 .1650 .1049 .1073 .0961
21) BRIM .1967 .2390 .2378 .2182 .2156 .1647 .1368 .1388 .1295
22) CAV P-1 .1951 .2388 .2384 .2149 .2154 .1635 .1369 .1390 .1296 23) CAV P-2 .1962 .2374 .2379 .2170 .2155 .1642 .1348 .1370 .1289 24) DIMI .1799 .2172 .2614 .2396 .2212 .1475 .1028 .1051 .0999
25) MIN P-1 .1960 .1775 .2192 .2192 .2085 .1243 .0769 .0785 .0765 26) P. N. SP. .3982 .4009 .4009 .4637 .4615 .4616 .5462 .5466 .5447
27) P. OBSC .3974 .4000 .4000 .4630 .4609 .4609 .5457 .5461 .5443




19 20 21 22 23 24 25 26


20) AUST .0941
21) BRIM .1263 .0867
22) CAV P-1 .1264 .0868 .0000
23) CAV P-2 .1253 .0861 .0004 .0004 24) DIMI .0959 .0656 .1006 .1008 .0981
25) MIN P-1 .0737 .0706 .1051 .1052 .1048 .0800 26) P. N. SP. .5471 .5481 .5481 .5480 .5474 .5477 .5468 27) P. OBSC .5467 .5474 .5474 .5474 .5468 .5472 .5463 .0921















DISCUSSION


Percent Polymorphism


Percentages of loci polymorphic have been used as a

measure of genetic variability in organisms studied electrophoretically. Of the various criteria that have been used to define polymorphic loci, the two most commonly employed have been to declare a system polymorphic if the frequency of the most common allele is equal to or less than 0.99 or

0.95. The percentage of all loci having more than one allele regardless of allele frequency is called the percent loci variable. Gorman and Kim (1976) found the percent loci variable averaged 35.5% for 15 populations of Anolis lizards. Three species of Bathygobius (inshore gobies) averaged 25.6% variable loci (Gorman et al., 1976). Hubby and Lewontin (1966) found 43% of the loci variable in Drosophila pseudoobscura. The 23 populations of Gomphus averaged 11.3%. The highest percentage of variable loci (27.3%) was found in Gomphus abbreviatus, G. n. sp. and G. exilis. By these comparisons Gomphus appear to be less variable than other organisms. However, Nei et al. (1975) point out that the percent loci variable, or polymorphic, is strongly dependent on sample size and therefore is not as good a measure of genetic variability as heterozygosity.


66




67


Heterozygosity


In random mating populations the average heterozygosity estimate is related to the proportion of an individual's loci that could be expected to be in a heterozygous condition. Even in nonrandom mating populations such as selffertilizing plants heterozygosity is a useful measure of genetic variability. In nonrandom mating populations the heterozygosity does not have a relationship to the frequency of heterozygotes. For this reason, Nei (1975) proposes substituting the term gene diversity for heterozygosity. As previously discussed, electrophoresis only detects about a third of the possible variation in the amino acid sequence of proteins. Therefore, the heterozygosity estimates presented are an underestimate of the total variability present in the genome.

Selander (1975) summarizes heterozygosities from many studies. In Drosophila, 28 species had an H of 0.150, 4 other unspecified insect species averaged 0.151. Nine species of marine invertebrates averaged 0.147. Vertebrates tend to exhibit lower levels of heterozygosity. Twenty-six rodent species averaged 0.054. Webster et al., (1973) found 4 species of Anolis had a mean H of 0.0285, including one species with no variability. Berlocher (1976) found 16 species of Rhagoletis had a mean H of 0.1053. The 23 taxa of Gomphus averaged 0.0221, considerably below the heterozygosities of most other organisms that have been studied, especially other insects. Although the H for the 9 species




68


of libellulids (0.0491) was twice the value for Gomphus it is still less than half as variable as the other invertebrates cited above. The low levels of variability in dragonflies may be attributable to their restricted habitat preferences, predatory nature, relatively small population sizes, and the possibility that they have undergone frequent genetic bottlenecks. Conversely, the migratory ability of most adult dragonflies would tend to offset, to some degree, the factors decreasing heterozygosity. The greater variability in the libellulids than in Gomphus may be correlated with a wider range of habitat preferences and generally larger population sizes.

Nei (1975) presents the following equation for estimating effective population size from heterozygosity: H = 4N V
e
Where H average heterozygosity

N = effective population size

V = mutation rate.

Kimura and Ohta (1971) suggest an average mutation rate of 10~7 per year for most proteins. Using this value and the H of G. abbreviatus, and solving for Ne, the effective population size is estimated to be 41,000. For G. pallidus the estimate is 129,250, for the average gomphine 54,500. The effective population size refers to the size of the panmictic breeding unit. It represents some quantity between the local deme and the total species population. The greater the gene flow between demes the closer the N will
e





69


be to the total species population. From m, personal field experience, these N estimates seem far larger than the size of local populations, indicating gene flow is probably significant.

Nei et al. (1977), and Chakraborty and Nei (1975) show quantitatively that the occurence of genetic bottlenecks can reduce levels of heterozygosity. The degree of reduction is related to the severity and frequency of bottlenecks, mutation rate and rate of population growth after bottleneck. The number of generations required for return to equilibrium is roughly the reciprocal of the mutation rate. Even with a mutation rate of only 10 6 that would mean a recovery period of a million years. Many Gomphus are obligatory stream dweLlers and presumably have been that way far back into their evolutionary history. The recent Pleistocene glaciations with their drastic effects on climate and sea level fluctuations probably had a disturbing effect on the restricted habitats of the Gomphidae. Considering the long period that would be needed to reach equilibrium, the low H observed today may be the sum effect of numerous glaciations and current small population size. If species could be sorted by susceptibility to the effects of glaciation future studies might be able to determine if in fact glaciation is responsible for the low levels of heterozygosity.






- 0


Genetic Distance and Proposed Phylogeny


Electrophoretic data can be used for both population genetics and systematics but the criteria required by the two disciplines can be quite different. Population genetics is concerned with the population dynamics within species, and large sample sizes are important to accurately estimate and follow small differences in gene frequencies. Systematists are interested in the evolutionary relationships between species and in the establishment and maintenance of a phylogenetically based classification system. Knowledge of the phylogeny within a related group can add considerable depth to the evaluation of comparative behavior studies. In population genetics it may sometimes be essential to confirm patterns of inheritance by breeding experiments. This is not practical with the majority of Odonata.

The number of specimens required to establish the order of species within a phylogeny depends on the number of loci examined, level of heterozygosity and the genetic distance involved (Nei, 1978). The larger the distance between species the fewer specimens required. In organisms with high levels of heterozygosity large sample sizes would be necessary for construction of a dendrogram. Gomphus, with its very low heterozygosities should require minimum sample sizes. Gorman and Renzi (1978), working with Anolis roquet and Anolis bimaculatus species groups, demonstrated that the correct topology of a dendrogram could be achieved with even single individuals of each species.




/ I


The phylogeny proposed for all genera studied is presented in Figure 23. The general outline of the family groups within the dendrogram is complementary with the current taxonomic interpretation. All family groups clustered cleanly, without overlap. The Macromiidae (represented by Didymops) clustered with the five genera of Libellulidae at a distance of 2.111. These 2 families then clustered with the Corduliidae (represented by Epitheca) at 2.576. Until recently, these 3 taxa had been considered only subfamilies of Libellulidae (Needham and Westfall, 1955). Gloyd (1959) elevated the Macromia group (including Didymops) to family status. Corduliidae has also since been recognized as a family (Walker and Corbet, 1975, among others). The Gomphidae and Aeshnidae cluster at 2.859. Fraser (1957), in his reclassification of the order placed the Aeshnidae and Gomphidae together in the superfamily Aeshnoidea and the Libellulidae and Corduliidae (Fraser considered flacromiidae as part of the Corduliidae) in the superfamily Libelluloidea. This division is strongly supported by the clustering of the 2 superfamilies at 4.443. The ordering of the 3 families in the Libelluloidea may not be accurately depicted in the dendrogram. Gloyd (1959) states, with good evidence, that the corduliids and libellulids are more closely related to each other than to the macromiids. The relationship is just the reverse in the dendrogram. At such a large distance (2.1 - 2.5), a more extensive survey, using more genera, would be required for resolution of this question.

























Figure 23. Dendrogram of the proposed phylogeny of 14 genera of Qdonata based
on genetic distance estimates from electrophoretic data.





1


0


4


3


2


-I


GOMPHUS

PROGOMPHUS GOMPHAESCHNA EPIAESCHNA CORYPHAESCHNA BASIAESCHNA ANAX EPITHECA LIBELLULA

PLATHEMIS ORTHEMIS LADONA

PACHYDIPLAX DIDYMOPS


7-4




74


Gomphus and Progomphus cluster at 1.783. Progomphus

is placed in the subfamily Gomphoidinae while Gomphus is in Gomphinae based on venational differences (Fraser, 1957). The larvae of Progomphus and Gomphus are also very different. Therefore, the large genetic distance between them is not surprising.

It is interesting to note that the 5 genera of aeshnids cluster at an average distance of 0.889, almost exactly the same as the average of the 5 genera of libellulids (0.893). This lends support to maintaining Plathemis and Ladona as valid genera rather than lumping them with Libellula as suggested by Ris (1910) and Kennedy (1922). Some current workers such as Walker and Corbet (1975) and Paulson and Garrison (1977) have adopted the usage of Ris and Kennedy. Needham and Westfall (1955) recognized both Plathemis and Ladona. Libellula depressa is the type species for the genus and is considered, on morphological grounds, to be most closely related to Plathemis lydia (Walker and Corbet, 1975). Any future attempt to resolve this question should include Libellula depressa. Figure 24 contains more details of the proposed phylogeny for the species of Corduliidae and Libellulidae surveyed. Orthemis ferruginea, placed by Fraser (1957) in the Libe'llulinae, clusters with Pachydiplax longipennis (subfamily Sympetrinae) before clustering with Libellula. The problems associated with subfamilies of Libellulidae and the resolution of species groups within Libellula could easily provide enough material
























Figure 24. Dendrogram of the proposed phylogeny of Corduliidae and Libellulidae
based on genetic distance estimates from electrophoretic data.


















rz


H-


E. CYNOSURA E. SEPIA E. STELLA E. COSTALIS L. INCESTA L. FLAVIDA L. AURIPENNIS L. NEEDHAMI L. SEMIFASCIATA P. LYDIA
0. FERRUGINEA L. DEPLANATA P. LONGIPENNIS


1.2 1.0 0.8 0.6 0.4 0.2 0


-l


I


II


-A


2 1.0 0.8


0.6 0 .4 0.2


i





77


for one or more additional dissertation research projects. I do not intend to draw any firm conclusions on the problems in Libellulidae. They were included only for comparison with the gomphines and for their obvious heuristic value. Many of the Libellulidae were represented by single specimens, a definitive study would require larger samples and the inclusion of more taxa, especially annectent forms.

The 4 species of Epitheca clustered at a very low distance level (0.0461). The validity of these species was demonstrated by Tennessen (1973 and 1977). The close relationship points out that species are defined in terms of reproductive isolation rather than genetic distance. However, electrophoretic analysis can prove quite helpful in delineation of new species (Ayala and Powell, 1972 and Berlocher, 1976).

The phylogeny proposed for Gomphus, based on the electrophoretic analysis of 22 loci, is presented in Figure 25. The subgeneric relationships are summarized in Figure 26. The branch points between groups of species in the dendrogram are obtained from the average of the separate pairwise distances of the species within the groups, therefore, the distance of the branch point may not match the distance between a particular pair of species.

All described species and one undescribed species of the subgenus Hylogomphus were represented in this study. G. parvidens proved to be most closely related (D=0.0438) to the new species. G. viridifrons and G. apomyius formed
























Figure 25. Dendrogram of the proposed phylogeny of Gomphus based on genetic
distance estimates from electrophoretic data.






















I








I


0.8 0.6 0.4 0.2 0


N. SP. P-I N. SP. P-2 N.SP. P-3 PARVIDENS
VIRIDIFRONS APOMYIUS ABBREVIATUS
BRIEVIS LINEATIFRONS DILATATUS
BRIMLEYI CAVILLARIS P-1 CAVILLARIS P-2 EXILIS P- 1 EXILIS P - 2 HODGESI LIVIDUS P- I LIVIDUS P - 2 AUSTRALIS DIMINUTUS MINUTUS P-1 MINUTUS P-2 SPICATUS
TOWNESI PLAGIATUS POTULENTUS
FURCIFER PALLIDUS
VILLOSIPES P. N. SP. P. OBSCURUS


11


I









-


0.8 0.6


0.4 0.2


0
























Figure 26. Summary of the subgeneric relationships in the proposed phylogeny
of Gomphus based on genetic distance estimates from electrophoretic
data.







HYLOGOMPHUS


GOMPHURUS GOMPHUS (in part)


GOMPHUS (in part) GOMPHUS (in part) STYLURUS


ARIGOMPHUS


0.8 0:6


0.4 0.2


mm'


I-


U


6





82


a tight cluster with each other at 0.0701. G. abbreviatus clustered with the first 4 species at 0.1556 and G. brevis joined at 0.2148, completing the subgenus Hylogomphus.

G. lineatifrons and G. dilatatus, the only representatives of the subgenus Gomphurus, clustered at 0.0868. This subgenus contains some of the largest and strongest flying Gomphus, their under representation in this study is partially due to the difficulty associated with their collection.

The next 3 populations were virtually identical at all loci. That G. brimleyi might be only a synonym of G. cavillaris has been suspected (Westfall 1974). He indicated they were structurally inseparable and that the only difference was that G. brimleyi was more darkly colored. Based on both the striking enzymatic and structural equality G. brimleyi should definitely be considered synonymous with G. cavillaris.

The 2 populations of G. exilis were identical at all loci, they were closely related to G. hodgesi (D=0.0309). Specimens of G. lividus from Florida, at the edge of their range, were very similar to specimens from central New York (D=0.0019).

The clustering of the 2 populations of G. minutus with G. diminutus presents some conceptual problems. G. minutus from NW Florida were, according to the enzymatic analysis, more closely related to G. diminutus (also from NW Florida) than they were to G. minutus from Gainesville, Florida.









G. minutus and G. diminutus can easily be told apart on the structure of the male abdominal appendages and other morphological characters. Of 3 possible explanations, introgression seems the least likely since intermediates between the 2 species have never been reported and also because G. diminutus has never been collected in proximity to the G. minutus population. The second possibility is that G. minutus may be a complex of sibling species, one of which is closely related to G. diminutus. Although this hypothesis seems unlikely and has no other supporting evidence, it cannot be disproven without further study. The third and most likely alternative is that the relationship is not phylogenetically accurate. The population of G. minutus from NW Florida is represented by only 2 specimens, the peninsular population by 28. The 2 G. minutus from P-1 were both very teneral, collected and frozen within a few hours of emergence. All other specimens in the study were fully matured adults. The teneral specimens may not have had their entire complement of adult enzymes fully functional. One of the findings of Anderson et al. (1970) was that protein composition changed with transition from larva to adult. The distance between G. diminutus and the larger, mature population (P-2) from Gainesville is probably closer to the actual relationship between the 2 species. G. australis clusters with G. minutus and G. diminutus at 0.1362.

The clustering of the subgenera is summarized in Figure 26. Species in the subgenus Gomphus did not cluster





84


cleanly. The G. cavillaris populations (including brimleyi) joined Gomphurus before clustering with the other Gomphus. The Gomphurus - G. cavillaris group then clustered with the Hylogomphus at D=0.2418. This assemblage then clustered with most of the other species in the subgenus Gomphus at

0.2803. G. spicatus, a somewhat abberent form, is the last species of Gomphus to cluster (0.3757). Needham (1948) wrote that the nymph of spicatus is somewhat intermediate, lacking a lateral spine on abdominal segment 6, a fact that once lead him to erroneously place spicatus in Arigomphus.

Of the 5 subgenera treated in Needham and Westfall

(1955),Stylurus and Arigomphus are clearly separated from the other 3. In Stylurus, G. townesi clusters with G. plagiatus at 0.2519, and they cluster with G. potulentus at 0.3487. In Arigomphus, G. pallidus and villosipes cluster at 0.3486, and they cluster at 0.4935 with G. furcifer.

Nei (1975) summarizes estimates of genetic distance

between taxa of different rank. The range of interpopulational distances (local races and subspecies) in a wide variety of vertebrates and invertebrates was 0.000- 0.351. In Gomphus the range was 0.0000 - 0.0191 (G. minutus P-1 and P-2 were excluded for the reasons discussed above). The range for sibling species in Drosophila (Hubby and Throckmorton, 1968) was 0.18 - 1.54. In Gomphus, G. lineatifrons and G. dilatatus (D=0.0868), and G. parvidens and G. n. sp. (D=0.0438) can be considered sibling species. E. cynosura,








E. stella and E. costalis are all sibling species. The largest distance between any 2 of them was 0.0461. Some non-sibling species pairs of Drosophila have distances as large as 2.54 (Hubby and Throckmorton, 1968). This is approximately the level at which families of dragonflies associate. The range for generic clustering for all dragonflies was 0.5023 - 1.783. Hence Gomphus seems to exhibit less genetic distance at all taxonomic levels. The question then arises as to what ranking should be assigned Arigomphus and Stylurus, the 2 groups clearly differentiated from the rest of Gomphus by the analysis of proteins. The branch point separating Arigomphus from the rest of Gomphus was

0.8083, similar to the levels attained by other genera in the order. The species of Arigomphus also seem to be a morphologically homogeneous group with similar habits and ecological requirements. I feel Arigomphus should be elevated to generic rank as originally suggested by Needham (1948).

The situation concerning Stylurus is not as easily resolved. Stylurus separated from the other species of Gomphus at 0.6612. This group is not, in my opinion, as morphologically or behaviorally distinct as Arigomphus. Walker (1958) also points out that the European G. flavipes and G. ubadschii may be intermediate between Stylurus and the rest of Gomphus. It is probably best to reserve judgement until future electrophoretic studies can include these species and more of the other North American species in Stylurus that were not represented in this study.




86


Faster evolving enzyme systems with higher heterozygosities such as EST-1 tend to contribute more to the estimate of genetic distance than monomorphic systems. Distance criteria used to assess taxonomic rank are strongly dependent on the loci included in a study. Thus a generic level distance from one study is not likely to be equivalent to the same measure from another study using a different group of loci. The solution to this problem is the inclusion of a small number of standard reference species in all studies. One of the primary objectives of future studies in dragonflies should be to increase the resolution of enzymes with high heterozygosities, which would result in an increase in the accuracy of proposed phylogenies.

Suggestions for future studies include determination of the generic limits in Libellula, further delineation of subgenera in Gomphus and clarification of the subfamilies of the Libellulidae. In the Zygoptera, studies of Enallagma and Argia would be of interest.


Neutralists Versus Selectionists


Considerable controversy revolves around the mechanisms

responsible for maintenance of polymorphic enzymes in natural populations. Johnson (1974), Kojima and Tobari (1969) and others represent the selectionists. They feel most or all enzyme polymorphisms are maintained by some form of Darwinian selection such as heterosis or frequency dependent selection. The neutralists, lead by King and Jukes (1969), Kimura and








Ohta (1974) and Nei (1975) feel that the majority of the observed variation is selectively neutral and the result of an equilibrium between mutation and genetic drift. This mutation equilibrium hypothesis does not preclude some small percentage of allozymes being maintained by selection, making it very difficult to disprove. Nei (1975) presents a convincing compilation of arguments favoring the mutation equilibrium hypothesis.

One of the most interesting correlates of the mutation equilibrium theory is that a given protein will accumulate variation at a constant rate through evolutionary time, implying the existence of a rough protein clock. Aside from the problem of calibration of the molecular clock the question of neutrality versus selection has little effect on the main objective of this study, the computation of genetic distances and construction of phylogenies.


Evolutionary Clocks


Nei (1972) proposed that the genetic distance might be proportional to time and that it could be used as a rough, statistical clock for estimating time since divergence. For an ideal molecular clock, all proteins would have to be evolving at a constant rate. This is certainly not the case (Sarich, 1977). Each protein is in itself a miniature timepiece. To achieve a reasonable estimate, a large number of loci are sampled with the hope that the different rates will average out. As much as possible the proteins to be analyzed




83


should be a representative sample of the genome, selected without reward to evolutionary rates. In practice, in a preliminary study such as this, that usually means all systems that resolve well enough to be scored.

Calibration settings of the clock have varied considerably. Nei (1975) sets 1 distance unit equal to 5 million years. Maxson and Wilson (1974) suggest a value of 18 million years estimated from an electrophoretic distanceimmunologic distance correlation. Gorman et al. (1976) hope to use the rise of the Panamanian isthmus to calibrate the clock. They are studying populations of fish from both the Atlantic and Pacific coasts that were apparently separated by the rise of the land bridge. The first group of species, on which they have reported, appear to be accumulating genetic distance at a rate consistent with the 18 million year calibration factor.

Tillyard (1917) indicated the Gomphidae and Aeshnidae existed as distinct groups all the way back into the early Cretaceous, a span of approximately 130 million years. The genetic distance for the split of Gomphidae and Aeshnidae was 2.859, indicating that 1 distance unit equals 45.5 million years. This value seems very large and may be due to an underestimation of the distance between the families. At such large genetic distances (greater than 2-3) there is an increased probability that back mutations will reverse previously accumulated differences, resulting in distance values biased towards similarity.















SUMMARY


Protein variation in the dragonfly genus Gomphus was studied using starch gel electrophoresis. A phylogeny was proposed for 23 species using genetic distance estimates derived from analysis of 22 genetic loci. An additional 23 species of Odonata from a wide variety of families were analyzed for comparison with Gomphus.

Average heterozygosities (11) for 23 species of Gomphus ranged from 0.0000 - 0.0852, with a grand mean of 0.0221. Six species of Gomphus had no apparent variability. Nine species in Libellulidae had a mean H of 0.0491. These H values are much lower than most other insects that have been investigated.

The range of genetic distance levels for local populations was 0.0000 - 0.0191, for sibling species 0.0438

0.0868 and for genera 0.5023 - 1.783. In general the proposed phylogeny agreed rather well with the existing classification of Odonata as established from conventional morphological studies. On the basis of genetic distance Arigomphus was raised from subgeneric to generic rank. Gomphus brimleyi, identical at all 22 loci with G. cavillaris, was synonymized.





90


Preliminary evidence based on small sample sizes indicates Ladona and Plathemis should not be Jumped into the genus Libellula.















APPENDIX A


Locality Data


Arigomphus furcifer New York, Tompkins Co., Ithaca, Conservation Pond at Junction of Rt. 366 and Upper Creek Rd.,

14 June 1976.

A. pallidus Florida, Alachua Co., Gainesville, 8, 9 and 19

April 1977.

A. villosipes New York, Tompkins Co., Ithaca, Conservation

Pond at Junction of Rt. 366 and Upper Creek Rd., 14

June 1976.

Gomphus dilatatus Florida, Liberty Co., Sweetwater Creek at

S-270, 10 May 1976.

G. lineatifrons West Virginia, Hampshire Co., North River at

Rt. 45, 6 and 7 June 1976.

G. australis Florida, Wakulla Co., Piggot Pond near Medart,

26 March 1977.

G. brimleyi Florida, Wakulla Co., Piggot Pond near Medart,

26 March 1977.

G. cavillaris P-1 Florida, Bradford Co., Sampson Lake,

27 March 1977.

P-2 Florida, Alachua Co., Melrose, Santa Fe

Lake, 2 April 1977.


91




92


G. diminutus Florida, Santa Rosa Co., BicT Coldwater Creek

at S-191, 30 March 1977.

G. exilis P-1 New Hampshire, Grafton Co., Livermore, Church

Pond 1 Km N of the Kancamangus Highway, 12 June 1976.

P-2 New Jersey, Ocean Co., Friendship Branch at

U.S. 70, 9 June 1976.

G. hodgesi Florida, Santa Rosa Co., Big Coldwater Creek at

S-191, 30 March 1977.

plus: Florida, Santa Rosa Co., Pond Creek at S-191,

7 May 1976.

G. lividus P-1 Florida, Liberty Co., Crooked Creek at S-270,

6 May 1976.

plus: Florida Liberty Co., Sweetwater Creek at S-270,

10 May 1976.

P-2 New York, Tompkins Co., Cascadilla Creek at

Turkey Hollow Rd., 14 June 1976.

G. minutus P-1 Florida, Alachua Co., Gainesville, Newnans

Lake, 25 March 1977, 1 and 8 April 1977.

P-2 Florida, Taylor Co., Econfina River at

Rt. 98, 26 March 1977.

G. spicatus New Hampshire, Grafton Co., Livermore, Church

Pond 1 Km N or the Kancamangus Highway, 12 June 1976.

G. abbreviatus West Virginia, Hampshire Co., North River at

Rt. 45, 6 and 7 June 1976.

G. apomyius Mississippi, Forrest Co., Chaney Creek, 1 Km E

of Brooklyn, 8 and 9 May 1976.





93


G. brevis Massachusetts, Middlesex Co., Squannacook River

at Groton Conservation Area, 11 and 13 June 1976.

G. parvidens North Carolina, Moore Co., Lakeview, Mill Creek,

2 June 1976.

G. viridifrons West Virginia, Hampshire Co., North River at

Rt. 45, 6 and 7 June 1976.

G. n. sp. P-1 Florida, Calhoun Co., Juniper Creek at Hwy

20, 6, 10 and 25 May 1976.

P-2 Florida, Santa Rosa Co., Pond Creek at S-191,

7 and 25 May 1976.

P-3 Florida, Liberty Co., Crooked Creek at S-270,

6 May 1976.

plus: Florida, Liberty Co., Sweetwater Creek at S-270,

10 May 1976.

G. plagiatus Florida, Santa Rosa Co., Blackwater River at

Bryant Bridge, 25 and 26 September 1976.

G. potulentus Florida, Santa Rosa Co., Pond Creek at S-191,

25 May 1976.

G. townesi Florida, Santa Rosa Co., Blackwater River at

Bryant Bridge, 25 and 26 September 1976.

Progomphus obscurus Florida, Calhoun Co., Juniper Creek at

Hwy 20, 25 May 1976.

P. n. sp. Florida, Calhoun Co., Juniper Creek at Hwy 20,

25 May 1976.

Anax junius Florida, Putnam Co., West Lake, 0.5 Km N of

Hwy 20, 14 March 1977.





94


Basiaeschna janata Florida, Santa Rosa Co., Big Coldwater

Creek at S-191, 30 March 1977.

Coryphaeschna ingens Florida, Alachua Co., Melrose, Santa Fe

Lake, 2 April 1977.

Epiaeschna heros Florida, Alachua Co., Gainesville, Newnans

Lake, 25 March 1977.

Gomphaeschna antilope Florida, Liberty Co., Crooked Creek at

S-270, 6 May 1976.

G. furcillata P-1 Florida, Wakulla Co., Piggot Pond near

Medart, 26 March 1977.

plus: Florida, Santa Rosa Co., Big Coldwater Creek at

S-191, 30 March 1977.

P-2 Florida, Alachua Co., Gainesville, Newnans

Lake, 1 April 1977.

Didymops floridensis P-1 Florida, Putnam Co., West Lake,

0.5 Km N of Hwy 20, 14 March 1977.

plus: Florida, Alachua Co., Melrose, Santa Fe Lake,

2 April 1977.

P-2 Florida, Wakulla Co., Piggot Pond near Medart, 26 March 1977.

D. transversa Florida, Wakulla Co., Piggot Pond near Medart,

26 March 1977.

Epitheca costalis Florida, Wakulla Co., Piggot Pond near

Medart, 26 March 1977.

plus: Florida, Santa Rosa Co., Big Coldwater Creek at

S-191, 30 March 1977.




Full Text

PAGE 1

PROTEIN VARIATION IN GOMPHUS (ODONATA: GOMPHIDAE) By KENNETH WILLIAM KNOPF A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1977

PAGE 2

ACKNOWLEDGEMENTS I sincerely thank the members of my graduate committee, Dr. Dale H. Habeck, Chairman, Dr. Minter J. Westfall, Jr., Dr. Lewis Berner, and Dr. Milton Huettel for their assistance and guidance during this study. I give special thanks to Dr. Habeck who has provided excellent facilities and friendly encouragement throughout my graduate study. Sincere appreciation is extended to Dr. Huettel for his valuable advice and the use of his laboratory which has made this study possible. I thank Ms. Candy Woodburn and Ms. Winifred Gaddis, of Dr. Huettel 's laboratory, who were responsible for much of my instruction in the technical aspects of electrophoresis. A special debt of gratitude is due Dr. Westfall, who was responsible for kindling my interest in Odonata. The most important acknowledgement to be made is to my wife, Suanne, whom I thank for providing technical, financial and emotional support. I thank Dr. and Mrs. William L. Peters, Florida A & M University, Tallahassee, for their warm hospitality and for the use of the University facilities in the Blackwater River State Forest. I gratefully acknowledge the assistance of Dr. Paul Fuerst, of the University of Texas at Houston, with computer programming of allele frequency data. ii

PAGE 3

I also express my appreciation to the Florida Department of Natural Resources for financial support during my graduate study. iii

PAGE 4

TABLE OF CONTENTS Page ACKNOWLEDGEMENTS ii ABSTRACT V INTRODUCTION 1 METHODS AND MATERIALS 5 Biochemical Background and Terminology 5 Overview of General Procedure 8 Species Studied 9 Enzyme and Buffer Systems 10 Preparation of Gels 13 Preparation of Specimens 15 Running of Gels 16 Staining, Fixation and Storage 19 Scoring of Gels and Data Processing 20 RESULTS 2 3 DISCUSSION 66 Percent Polymorphism 66 Heterozygosity 67 Genetic Distance and Proposed Phylogeny 70 Neutralists Versus Selectionists 86 Evolutionary Clocks 87 SUMMARY 8 9 APPENDICES A LOCALITY DATA B BUFFER AND STAIN FORMULATIONS LITERATURE CITED BIOGRAPHICAL SKETCH iv 91 97 103 107

PAGE 5

Abstract of Dissertation Presented to the Graduate Council of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy PROTEIN VARIATION IN GOMPHUS (ODONATA: GOMPHIDAE ) By Kenneth William Knopf December 1977 Chairman: Dale H. Habeck Major Department: Entomology and Hematology Protein variation in the dragonfly genus Gomphus was studied using starch gel electrophoresis. A phylogeny was proposed for 23 species using genetic distance estimates derived from analysis of 22 genetic loci. An additional 23 species of Odonata from a wide variety of families were analyzed for comparison with Gomphus . Average heterozygosities (H) for 23 species of Gomphus ranged from 0.0000 0.0852, with a grand mean of 0.0221. Six species of Gomphus had no apparent variability. Nine species in Libellulidae had a mean H of 0.0491. These H values are much lower than most other insects that have been investigated . The range of genetic distance levels for local populations was 0.0000 0.0191, for sibling species 0.0438 0.0868 and for genera 0.5023 1.783. In general the proposed phyv

PAGE 6

logeny agreed rather well with the existing classification of Odonata as established from conventional morphological studies. On the basis of genetic distance Arigomphus was raised from subgeneric to generic rank. Gomphus brimleyi Muttkowski, identical at all 22 loci with G. cavillaris Needham, was synonymized. Chairman

PAGE 7

INTRODUCTION The status of the subgeneric rankings within the genus Gomphus have been in dispute for some time. Needham (1901) first organized the known North American species of Gomphus into four subgenera: Ar igomphus , Gomphus , Gomphurus and Stylurus . Needham (1948) proposed that Arigomphus and Stylurus be elevated to generic status. In 1951, he introduced a fifth group, Hylogomphus , comprised of 4 species previously included in the subgenus Gomphus . In 1955, Needham and Westfall treated all five groups as subgenera. Some authors accepted the elevation of Stylurus even before Needham' s 1948 paper. Williamson (1932) described 2 new species and placed them in the genus Stylurus . Walker (1958) followed the subgeneric rankings of Needham and Westfall (1955) but stated he had found Arigomphus , Stylurus and Gomphurus to be clearly defined groups based on study of the penis, vulvar lamina and hamules . Fraser (1957) recognized ^ Stylurus and Gomphurus as valid genera but not Arigomphus . The analysis of electrophoretically demons tratable protein variation has proven to be one of the most important tools of systematists and population geneticists in the study of evolution (Avise, 1974) . Much of the pioneering application of this technique to questions of evolutionary 1

PAGE 8

2 concern was conducted on Drosophila (Hubby and Lewontin, 1966, Ayala and Powell, 1972, and Ayala and Tracey, 1973). Since the late 1960 's a wide variety of organisms other than Drosophila have been subjected to electrophoretic scrutiny. Examples of typical studies, although by no means an inclusive list, include: Gorman and Kim (1976) proposed a detailed phylogeny of Anolis lizards based on 22 enzyme loci, Selander et al., (1971) studied the variation between subspecies of the mouse, Peromyscus polionotus , King and Wilson (1975) compared humans and chimpanzees and Gorman et al., (1976) analyzed variation in littoral gobies. The basic hypothesis underlying taxonomic analysis of protein variation is that over evolutionary time, due to the random nature of mutation, isolated populations will accumulate electrophoretically detectable changes in allele frequencies. Changes should accumulate at a rate proportional to the mutation rate. Thus, the longer the time since isolation, the greater the likelihood that populations can be differentiated by electrophoresis. The only work reported on dragonflies is that of Ander•s son et al. (1970). Hemolymph samples from larvae of Anax junius , Aeshna umbrosa and Libellula pulchella were compared using acrylamide gel disc electrophoresis and an AmidoSchwarz general protein stain. They reported clear separation of all three species and postulated that the technique might be valuable in describing phylogeny.

PAGE 9

3 There is considerable controversy concerning the mechanisms responsible for the maintenance of polymorphic variation in natural populations. Kojima and Tobari (1969) and Johnson (1974) proposed that natural selection is responsible for the observed variation. Kimura (1969) , Ohta (1972) and Nei (1975) are the main proponents of the neutral mutation hypothesis which says most of the molecular polymorphisms are maintained by the random walk of frequencies of selectively neutral alleles. The implications of this controversy for measures of genetic distance and estimates of time since divergence will be discussed later. The specific objectives of my research were: 1) To obtain heterozygosity estimates for members of the genus Gomphus and other selected odonates and compare them with each other and with published heterozygosity values for other organisms. 2) To construct a phylogeny of Gomphus from electrophoretic genetic distance data and compare that phylogeny with the existing classification derived from conventional morphological studies. Special emphasis was placed on resolution of the classification of the groups included in Gomphus . 3) To serve as a supplementary procedural guide for electrophoresis of dragonflies. The discussion will include suggestions for further research on other groups of Odonata where analysis of protein variation might significantly improve classification.

PAGE 10

4 Prospective researchers are encouraged to consult more extensive methods manuals such as Bush and Huettel (19 72) , Brewer (1970), Shaw and Koen (1968) and Harris and Hopkinson (1976) .

PAGE 11

METHODS AND MATERIALS Biochemical Background and Terminology Proteins are made up of long chains of amino acids, of which there are about 20 commonly occurring varieties. The majority of amino acids are electrostatically neutral. The acidic amino acids with negative charges are aspartate and glutamate. The basic amino acids lysine, arginine and sometimes histidine are positively charged (Stryer, 1975). The primary structure of proteins consists of the sequence of the amino acids which compose it. Secondary structure refers to patterns formed by hydrogen bonding between amino acid residues close to each other in the primary sequence. Tertiary structure refers to the complex folding of the chain to determine the final shape of the molecule. A protein is said to have a quaternary structure when two or more separate protein chains are closely associated. The associated subunits can be identical, products of different alleles at a polymorphic locus or products of different but related loci. The most widely known example of the third possibility is the hemoglobin molecule in man (Stryer, 1975). Hemoglobin has four separate chains and is called a tetramer; since its subunits are not identical it is further defined as a heterotetramer. Multimers (having more than one unit) with identical 5

PAGE 12

6 subunits are prefixed with homo-, such as a homodimer consisting of two equal subunits. Proteins with only one peptide chain are called monomers. Electrophoresis can be defined as the separation of differently charged particles in an electric field. Stryer (1975), in elaborating on the determinants of electrophoretic mobility, states: E (Z) v= ~! where: V = Velocity of migration E = Strength of electric field Z = Net charge on the protein f = frictional resistance The frictional resistance is a function of the size and shape of the molecule versus the pore size of the medium. The water soluble proteins in this study are almost all globular in shape. Since allelic differences coding for alteration of one of a few amino acids in the vast structure of the whole protein would not be expected to have much effect on the shape, the frictional resistance can be ignored. The strength of the electric field is roughly identical for all specimens in a single gel and therefore can also be ignored. The velocity can then be considered directly proportional to the charge on the molecule. A change of even a single charge unit can be detected. Often this may represent mutation of only a single amino acid. The substitution of

PAGE 13

7 neutral valine for negatively charged glutamate at the 6th position in the beta chain of hemoglobin A yields hemoglobin S. Hemoglobin S is responsible for sickle cell anemia. These 2 proteins, differing in only one amino acid, can be separated by electrophoresis (Stryer, 1975) . Since most amino acids in a protein are not charged i their mutation to another uncharged amino acid is not detectable unless the shape of the molecule is drastically altered. Nei (1975) estimates the probability of a mutation causing an electrophoretically detectable charge change at roughly 0.25 to 0.30 using both theoretical and empirical data. Therefore, assuming accumulation of mutations over time, the longer the time since divergence^ the higher the likelihood populations will be separable by electrophoresis. This relationship holds until very large genetic distances are reached. At distances greater than 2 or 3 the probability increases that charge differences will be cancelled out by additional mutations. Therefore, the distance between distantly related organisms will tend to be underestimated. Because of this phenomenon, differences between taxa should be stressed more than similarities. The terminology in this report will, for the most part, follow that of Prakash and Lewontin (1968) . The term isozyme is used to refer to 2 or more enzymes that can be stained simultaneously using the same procedure for both. They presumably have identical biochemical roles in the organism. Allozymes, mobility variants within a single locus, will most

PAGE 14

3 often just be called alleles. Allozymes are therefore also isozymes, but isozymes, if coded for by different genes, are not allozymes. King and Wilson (1975) have proposed calling allozymes "electromorphs" to emphasize that electrophoresis only reveals a portion of the genetic variation present. I have continued with the more widely used allelic terminoSpecimens to be electrophoresed were homogenized in separate drops of buffer. Each crude homogenate was then soaked up on one or more small pieces of chromatography paper. A horizontal slit was cut in the precast starch gel and the specimen papers inserted. A direct current was passed through the gel perpendicular to the slit for the papers. Proteins having a net positive charge migrate toward the negative cathode and those having a net negative charge toward the positively charged anode. Most enzyme s/ proteins are in the latter category. The length of time the current is applied is dependent on the consistency of the gel, buffer systems employed and the migration rate of the enzyme being studied. The gel was then trimmed, removed from its mold and sliced into thin sheets. These were stained for any of a large variety of specific enzyme systems by use of the appropriate histochemical technique. Gel slices were incubated in the stain solution at 37°C, with occasional agitation until the banding pattern developed. logy . Overview of General Procedure

PAGE 15

9 The gels were then preserved in fixative and refrigerated. After being photographed they were evaluated and scored. Species Studied A list of dragonfly species studied is given in Table 1. Where more than one population was studied, it is indicated by assignment of a population number. The taxonomic interpretation is close to that of Needham and Westfall (1955) . The non-gomphine species were included to help establish approximate genetic distance criteria for taxonomic levels within the order and to have a wider base of comparison for the analysis of variation within Gomphus . Appendix A contains collection data and localities for all species studied. All determinations were made by the author. Enzyme and Buffer Systems An annotated list of recipes for stains and buffers used in this study, along with a list of chemical abbreviations, are included in Appendix B. Table 2 lists the 22 enzymes, and their abbreviations, that were actually scored and represent the basis for all further computations. Table 3 lists the additional enzymes that were tried during preliminary screening. These were not scored for any of a variety of reasons such as faintness, poor resolution, lack of repeatability, or total lack of staining. Just the head and thorax of each specimen were used, the abdomens were saved as voucher specimens. Some of the enzyme systems not

PAGE 16

10 Table 1. List of dragonfly taxa. GOMPHIDAE Gomphus Subgenus Arigomphus f urcif er Hagen pallidus Rambur villosipes Selys Subgenus Gomphurus dilatatus Rambur lineatif rons Calvert Subgenus Gomphus australis Needham brimleyi Muttkowski cavillaris Needham diminutus Needham exilis Selys hodgesi Needham lividus Selys minutus Rambur spicatus Hagen Subgenus Hylogomphus abbreviatus Hagen a pomyius Donnelly brevis Hagen parvidens Currie viridif rons Hine sp. n. Subgenus Stylurus plagiatus Selys potulentus Needham townesi Gloyd Progomphus obscurus (Rambur) sp . n . AESHNIDAE Anax junius (Drury) AESHNIDAE continued Coryphaeschna ingens (Rambur) Epiaeschna heros (Fabricius) Gomphaeschna antilope (Hagen) f urcillata (Say) MACROMI IDAE Didymops f loridensis Davis transversa (Say) CORDULIIDAE Epithec a costalis (Selys) cynosur a (Say) sepia (Gloyd) Stella (Williamson) LIBELLULIDAE Ladona deplanata (Rambur) Libellula auripennis Burmeister f lavida Rambur incesta Hagen needhami Westfall semif asciata Burmeister Orthemis f erruginea (Fabricius) Pachydiplax longipennis (Burmeister) Plathemis lydia (Drury) Basiaeschna janata (Say)

PAGE 17

11 Table 2 . Names and abbreviations for the 22 enzymes scored in this study. Name Abbreviation rllUbUIlaLabc X IV r~* i H P h n q n h ^ ^~ ^ cp — 0 ACPH-2 t\\X til iy la l_ i\ Ilia oc ADK A 1 mho 1 nphvHrnnpn^ qp ADH t\ X \J\J X d o t: /A J_l J — I v_J l_ V_ X, LA O V -L EST-1 pcfpT-aep-2 EST-2 fienpral Profpin VJ \ — X X ^ — J_ -I— X 1 \ 1— v _i_ X X GEN PRO Glu tdma te Dehydrogenase GDH a—Gly ceropho spha te Dehy drogena se GPD Hexanol Dehydrogenase HEX-DH W fii y r"\ lr -i cp 11 Ej .A u nyu.xvJAyjJU.i_.yi.ciut: JJtriiycaxu i yt:Xiciot; xfaocicraue uenyurogenase j. T Pi I I 1 Isocitrate Dehydrogenase-2 IDH-2 Lactate Dehydrogenase LDH Leucine Amino Peptidase LAP Malic Enzyme ME Phosphoglucomutase PGM Phosphoglucose Isomerase PGI Tetrazolium Oxidase-1 TO-1 Tetrazolium Oxidase-2 TO-2

PAGE 18

Table 3 . Enzyme systems surveyed but not scored. Enzyme Staining? Aldehyde Oxidase no Alkaline Phosphatase yes Fumerase yes Galactose Dehydrogenase no Glucose-6-Phosphate Dehydrogenase yes Glutamate-Oxaloacetate Transaminase no Glyceraldehyde-3-Phosphate Dehydrogenase yes Leucine Amino Peptidase (old) no Malate Dehydrogenase yes Monoamine Oxidase yes Octanol Dehydrogenase yes Pepsinogen yes 6-Phosphogluconate Dehydrogenase no Sorbitol Dehydrogenase no Succinate Dehydrogenase no Triosephosphate Isomerase yes Tyrosinase no Xanthine Dehydrogenase no

PAGE 19

13 resolved may have been specific to abdominal tissues. In future studies where voucher specimens are not necessary some of these enzyme systems might prove useful. To save time for any future studies that might try to improve on the resolution, I have indicated which systems stained even though they could not be scored. Table 4 indicates the most common buffer-enzyme combinations employed, duration of the run, and the amperage or voltage to be maintained during the run. Preparation of Gels A mixture of 34.6 g Connaught hydrolysed potato starch (Fisher Scientific, New York, N.Y.), 11.4 g Electrostarch (Electros tarch Co., Madison, Wisconsin) and 20 g sucrose was dissolved in 400 ml of gel buffer in a 1000 ml flask. The quantities of starch and sucrose were empirically chosen. They can be varied along with the relative proportions of each in an attempt to maximize band resolution. The mixture was then heated on a stirring hot plate until it was uncomfortably hot to the touch. At this point the stir bar was removed and, with an asbestos glove, the flask was swirled vigorously over the open flame of a gas burner. As the gel heats it thickens and becomes translucent. Heating was continued until the gel just began to boil. A slight reduction in viscosity may be noticed before the boiling point is reached . The gel was then degassed for about 4 5 seconds using a Venturi-type water aspirator. The aspirator was connected

PAGE 20

14 Table 4. The most common enzyme-buffer combinations used in this study with the duration of run and voltage and amperage settings. I. Electrode buffer Borate Gel buffer Tris-HCl Run DurationSettingsEnzymeslh hours 250 V ACPH ADH TO 4 hours 250 V LDH GEN PRO II. Electrode buffer Borate Gel buffer Poulik Run durationUntil 7.5 to 8.0 cm brown zone migration SettingsEnzymes180 V PGI HBD LAP III. Electrode buffer Histidine-8 . 0 Gel buffer Histidine-8 . 0 Run durationSettingsEnzymesGPD ALD PGM IDH ME 4 *5 hours 5 0 mAmps 3h hours 50 mAmps EST ADK IV. Electrode buffer Tris Gel buffer Poulik Citrate Run durationSettingsEnzymes3h hours 5 0 mAmps GDH HEX-DH HEX

PAGE 21

15 directly to a 2 hole stopper, suction being applied by fitting the flask to the stopper and covering the second hole with a finger. The vacuum must be released by removing the finger from the hole rather than by shutting off the water. Safety glasses and a shield are a necessity since there is a real danger of implosions. The degassed gel was then poured quickly into the center of a 1.0 cm deep, 17.5 x 19.5 cm square, plexiglas mold. This same mold was used to hold the gel until it was sectioned and stained following the run. Small pieces of debris could only be removed from the gel while it was still molten. After the gels cooled to room temperature (about 1 hour), they were wrapped in clear plastic food wrap to retard drying. Gels stored for a day at room temperature give the most consistent results. Gels were refrigerated for just one or two hours prior to specimen application. Preparation of Specimens Specimens to be electrophoresed were killed by freezing in a cooler of dry ice while on field trips. On return to the laboratory they were transferred to an ultra low temperature freezer for storage. Temperatures well below freezing are required to prevent enzyme degradation and denaturation . The head and thorax of each frozen specimen was pulverized in a round-bottom, 2 ml, plastic vial using a Delrin R plastic rod. The abdomen was retained for confirmation of species identification. This whole operation was conducted

PAGE 22

16 on a bed of dry ice with prechilled tools. Special care was taken to avoid any heat transfer from the hands to the samples. All grinding rods were carefully wiped clean after each specimen was pulverized to prevent contamination. Since some of the enzymes to be surveyed may have been tissue specific, pulverization of specimens made a potentially more homogeneous sample. An additional advantage of pulverization is that only a small portion of the sample is required for a run; thus the same individual can be run repeatedly. This flexibility allows many side by side comparisons between taxa found, in hindsight, to be closely related. Running of Gels A clean, metal spatula, moistened with gel buffer, served to pick up an adequate subsample of the frozen, pulverized head and thorax without thawing the sample. A single pulverized specimen of Gomphus provided enough material for 5 runs. The subsamples were deposited in the separate holes of a prechilled Delrin plastic grinding block with a few drops of grinding buffer. A tray of crushed ice was used to keep the block chilled during the entire process. Individually assigned specimen numbers were recorded on protocol sheets for each grinding. The sample was ground (mixed) using a Delrin grinding rod chucked into a variable speed stirrer until a relatively smooth homogenate was obtained (about 10 to 20 seconds) . After mixing was complete for all individuals to be run, separate rectangles of 1.0 x 0.4 cm

PAGE 23

17 Whatman R No. 1 chromotography paper were used to soak up the homogenate. Up to 5 pieces of paper were used in each hole. A slit was cut in the chilled starch gel 5.5 cm from one end using a clean metal spatula and a ruler. The slit in the gel was eased open with the fingers and the lightly blotted sample papers inserted with a forceps. Twenty-four separate papers of 0.4 cm width could be put in each gel with approximately 0.3 cm between each paper and 1.5 cm on each end of the row to allow room for trimming the gel. When the slit was closed the gel was ready to be set up for the run . Details of the construction of the buffer boxes, power supplies and safety boxes used are described by Bush and Huettel (1972). Each plexiglas buffer box holds approximately 450 ml. Buffer boxes containing platinum wire electrodes were placed in a modified plastic food crisper and the remainder of the food crisper packed with crushed ice. The gel tray was placed on top of the buffer boxes with the slit in the gel parallel to the buffer boxes. Sponge cloths served as wicks to bring the buffer and current up into the gel. The gel was covered with clear plastic food wrap from a line parallel to the slit and 3.5 cm from the cathodal end of the gel to beyond the anodal end of the gel. The sponge cloth for the cathodal end of the gel was drawn up to about 1 cm from the slit in the gel. Thus the cathodal sponge overlapped the clear plastic wrap by 1 cm. The food

PAGE 24

18 wrap helped prevent drying and shrinkage during the run. The anodal sponge was brought up under the food wrap to a line 9.0 cm anodal of the slit. The gel and sponges were then covered with a glass plate that served both for electrical insulation (safety) and as a support for a pan of ice to keep the gel from overheating during the run. The power supplies were connected and set (Table 4) and the run began. Electrical resistance of the gels change during the course of the run so settings must be checked repeatedly. With the borate electrode Poulik gel buffer system, a brown zone could be seen where the consistence of the gel changed, presumably as the buffer of the electrode passed through the gel. With this system the run was terminated after a satisfactory migration distance was obtained (usually 7.5 to 8.0 cm) . On buffer systems without a visible front, the length of the run was determined empirically. Table 4 contains the durations of runs used with each specific buffer-stain system. After turning off the power, the gel in its tray was removed from the apparatus and trimmed with a clean metal spatula and a straight edge. A centimeter of gel was trimmed off and discarded from each side as well as off the cathodal end of the gel. A cut across the gel was made 9.0 cm anodally from the slit and the portion that was under the anodal sponge discarded. The remainder was the larger, approximately 16.5 x 9.0 cm front or anodal slice, and the approximately 4.5 x 16.5 cm back or cathodal slice. Before they were removed from the gel tray they were marked for future orientation. A notch was taken off the left side of

PAGE 25

19 each gel on the corner away from the slit. Most enzyme systems travel in the anodal direction. Only TO-1 and sometimes PGI were cathodal migrants in the dragonflies studied. Therefore, during the majority of the runs a back slice was not stained. However, during preliminary surveys, a back slice was always stained to ascertain which way the protein migrated. The trimmed gel was placed on a plexiglas slicing tray. Air bubbles were cleared from beneath the gel by pressing gently on the top of the gel while observing from below. The gel slicer was a modified hacksaw that holds a thin stainless steel wire instead of a blade. The slicer was brought through the gel in one smooth, continuous motion with the wire held firmly down against the 2 mm high guide rails of the slicing tray. The thin slice was removed for staining. Three and sometimes 4 slices could be obtained from each gel. After each cut the wire was carefully wiped clean to prevent small bits of gel from hardening on the wire and resulting in the following slices being very rough surfaced. Staining, Fixation and Storage Stain solutions were prepared as close as possible to the time of their use, since they often contain unstable cof actors and enzymes. The dry ingredients were weighed out while the gels were running and stored in the refrigerator in small flasks. Liquid components were added at the

PAGE 26

20 last possible minute. Gel slices were incubated (37°C) according to the appropriate recipe (Appendix B) until the banding patterns developed. The staining solution was then discarded and about 75 ml of fixative added. Gels changed from translucent to opaque as fixation progressed. The following day gels were removed from the fixative, photographed and wrapped in clear plastic food wrap (Saran R seems to work best) . Wrapped gels may be stored for more than a year in a refrigerator. Scoring was done as soon as possible, since a few of the systems fade with storage. Scoring of Gels and Data Processing As is necessary in all electrophoret ic studies where mating tests are impractical, the banding patterns on the gels were interpreted as direct gene products. In organisms such as Drosophila (Hubby and Lewontin, 1966) , Gryllus (Harrison, 1977) and Rhagoletis (Berlocher, 1976) where crossing is possible , most enzyme banding patterns tested by breeding experiments proved to be genetically based allelic variation. Dragonfly biochemistry is not likely to be grossly different from that of these organisms. The allelic form most common in Gomphus abbreviatus was taken as a standard reference allele. Faster and slower forms were compared by dividing their migration distances into that of the standard. By continually varying the order the specimens were put on the gel, side by side comparisons were made between most of the possible combinations of species studied.

PAGE 27

21 To estimate the degree of genetic similarity between populations the gene frequency data for each population were compared in a pair-wise manner to all other populations. The result is a matrix of numerical estimates of similarity for all possible combinations. The measure of similarity selected was Nei's (1978) estimate for the unbiased genetic distance since it compensates for a slight overestimate by the standard formula when applied to small sample sizes. A genetic distance (D) of zero indicates the proteins of that pair of populations were electrophoretically identical at all loci sampled. The estimate of the standard distance is given by D = -In (G X y/ /G X G ). The average population gene identities G x , Gy, and G xy are usually replaced by the observed sample frequencies J x , J y , and J xy , which are the averages of ? 2 £ x i i E Yi i ar >d £x^y^ over all loci, where is the gene frequency of a particular allele in population x and y. is the frequency of the same allele in population y. To obtain the unbiased estimate of D for populations X and Y where the number of individuals, n^ and n y can vary from locus to locus, G x and G y are estimated by the average over all loci of (2n x Zx i 2 l)/(2n x 1) and (2n y E Yi 2 l)/(2n y 1), respectively . Heterozygosity estimates were obtained with the formula 2 H = 1 £ x i , averaged over all loci, where is the frequency of the i-th allele.

PAGE 28

22 Copies of the computer programs used to analyze the allele frequency data in this study are available upon request from Dr. Masatoshi Nei, Center for Demographic and Population Genetics, University of Texas Health Science Center at Houston, P.O. Box 20334, Houston, Texas 77025. The programs have also been adapted to the IBM 360 computer at the University of Florida (NERDC) and are available through Dr. M. D. Huettel, Insect Attractants Behavior and Basic Biology Research Laboratory, USDA, Gainesville, Florida.

PAGE 29

RESULTS Observed allele frequencies for Gomphus and Progomphus are presented in Tables 5-9. A total of 110 alleles were resolved in the 22 loci surveyed in Gomphus . With the addition of Progomphus and the rest of the species studied in other families of Odonata, the total number of separate alleles was 304. Relative allele mobilities of Gomphus and Progomphus are given in Figures 1-22. Of the 22 loci scored, 10 (45.5%) were monomorphic in all species of Gomphus . Four of those 10 (TO-1, GDI! , ACPH-1 and ALD) exhibited a single monomorphic allele occurring in all species. Percent of loci polymorphic and percent loci variable for Gomphus and Progomphus are presented in Table 10. Table 11 presents heterozygosity estimates for all loci and the average heterozygosity (H) for each population of Gomphus and Progomphus . Average heterozygosities for Gomphus ranged from 0.0000 to 0.0852 and the mean for the 23 taxa was 0.0221. Six populations had no apparent variability. The heterozygosity data for the species of Odonata studied that were not in the family Gomphidae are summarized in Table 12. The non-gomphines were included just for comparison purposes and possible heuristic value. Many of 23

PAGE 30

2 4 these species were represented by very small samples. Therefore, the details of gene frequencies and relative allele mobilities are not presented. The average heterozygosity for P rogomphus obscurus was 0.0432 and for P. n. sp. 0.0270. In the non-gomphines, 9 species of libellulids had a mean H of 0.0491 (Table 12). This is twice as high as the average value in Gomphus . Ladona deplanata , a libellulid, had the highest H (0.1111) of all species surveyed. The matrix of unbiased genetic distance estimates for Gomphus and Progomphus is presented in Table 13. Table 14 contains a matrix of standard errors for the distances listed in Table 13. There is no formula for direct computation of the standard error of the unbiased distance so the standard error of Nei's standard distance is presented as a close estimate.

PAGE 31

25 Table 5 . Allele frequencies of Gomphu s abbre viatus, G. brevis , G. viridif rons , G. apomyiu s^ and 3~ populations of a new species of Gomphus. 2 Enzyme Allele 2 2 2 CO in 00 * *x) t) 1 1 if ro < H H C3 >v i 1 1 1 UJ H » O ro LJ U! |X3 < l — i LDH N 35 47 5 87 73 12 7 6 100 1.00 1.00 1 . 00 1.00 1. 00 1.00 1.00 1. 00 HEX N 35 47 5 87 73 12 7 6 100 1 . 00 1.00 1. 00 102 1 . 00 1. 00 1 . 00 1. 00 1.00 IDH-1 N 36 44 5 85 73 12 7 6 100 1.00 1. 00 1 . 00 1. 00 1.00 1. 00 1.00 1. 00 IDH-2 N 36 46 5 87 73 12 7 6 100 1. 00 1.00 1 . 00 1. 00 1 . 00 1.00 . 929 1. 00 121 . 071 PGM N 34 46 5 85 72 12 ~1 6 91 _____ 1.00 100 1.00 .989 1.00 .994 1.00 .857 1.00 107 ______ >143 _ 136 .011 .006 ADH N 35 47 5 87 75 12 7 6 89 . 083 100 1. 00 .809 1 . 00 .897 1.00 . 875 1. 00 1 . 00 138 .191 . 103 .042 TO-1 N 33 46 5 84 72 12 7 5 100 1.00 1.00 1 . 00 1.00 1.00 1.00 1.00 1 .00 TO2 N 31 45 5 81 68 12 7 5 35 1.00 1.00 1 .00 1.00 . 007 100 .993 1.00 1.00 1 . 00 GEN PRO N 36 46 5 87 75 12 7 6 100 1.00 1. 00 1 . 00 1. 00 1. 00 1. 00 1. 00 1 .00 ME N 32 , 43 5 80 71 8 5 4 53 . 028 100 1. 00 1. 00 1 . 00 1.00 .972 1.00 1. 00 1 . 00 GPD N 35 47 5 87 75 12 7 6 57 . 042 100 1. 00 1.00 1. . 00 1.00 1.00 .958 1. 00 1, .00 PGI N 24 29 1 54 60 8 5 4 45 . 063 47 . 008 61 . 017 .009 67 . 075 100 1.00 . 931 1. 00 . 963 . 917 . 937 1.00 1. 00 160 . 052 . 028

PAGE 32

2 6 Table 5 . Continued. 3 Enzyme Allele S3 • • GO in • * i > W < H > H tD W I 1 1 1 w RI o H M < 3 HEX-DH N 31 43 5 79 69 9 5 3 87 — ~ — — .029 100 1.00 1.00 1. 00 1 . 00 .971 1 . 00 1.00 .333 121 — — . 667 ADK N 14 9 1 24 23 4 3 2 100 1 . 00 1.00 1.00 1. 00 1.00 1. 00 1.00 1. 00 HBD N 11 12 1 24 13 7 5 4 67 . 045 — . 021 — 100 . 955 1.00 1. 00 .979 1.00 1.00 1. 00 1700 — 1. 00 GDH N 30 38 5 73 63 9 5 4 100 1.00 1. 00 1. 00 1.00 1.00 1.00 1. 00 1.00 ACPH-1 N 35 47 5 87 75 12 7 6 100 1. 00 1.00 1. 00 1.00 1. 00 1.00 1.00 1.00 ACPH-2 N 35 47 5 87 75 12 7 6 100 1.00 1.00 1.00 1. 00 1.00 1.00 1.00 1.00 LAP N 21 40 5 66 68 12 7 6 100 1.00 1. 00 1. 00 1.00 1. 00 1.00 1.00 1. 00 EST-1 N 34 49 5 88 73 11 7 6 70 . 015 . 006 84 . 071 . 083 90 . 912 . 959 . 300 . 914 92 . 073 . 031 . 700 . 075 93 . 045 95 . 007 . 955 . 857 . 667 98 . 072 100 . 010 . 005 . 973 104 .250 106 . 020 EST-2 N 36 46 5 87 74 12 7 6 54 . 022 . 012 70 . 014 .011 . 012 100 .986 .956 1.00 . 971 . 986 1.00 1.00 107 1.00 125 . 014 140 . 011 . 005 ALD N 26 31 1 68 72 8 5 4 100 1 . 00 1 . 00 1. 00 1.00 1.00 1.00 1. 00 1.00

PAGE 33

27 Table 6 . Allele frequencies of Gomphus parvidens , G. lineatif rons , G. dilatatus , G. f urcif er , G. pallidus , G. villosipes , G. townesi, and G. potulentus . Enzyme Allele J)t H o^'x)nt H lr l 2!G LDH N 1 10 1 10 24 1 3 10 80 1.00 1.00 1.00 100 1.00 1.00 1.00 1.00 1.00 HEX N 1 10 1 10 21 T 8 10~ 97 1.00 100 ______ 1>00 102 1.00 1.00 1.00 1.00 1.00 1.00 IDH-1 N 1 10 1 10 24 1 8 9 93 ______ 1.00 100 1.00 1.00 1.00 _____ 109 _______ 1>00 122 1.00 1.00 130 1.00 IDH-2 N 1 10 1 10 T4 I 8 9~ 60 .150 100 1. 00 1. 00 1. 00 . 350 1. 00 1. 00 1. 00 l.QQ PGM N 1 10 1 10 24 1 ~8 10~ 38 ______ i_ 00 .950 93 .100 1.00 100 1.00 1.00 1.00 102 _ .goo 107 1.00 _24 .05 0 ADH N 1 10 1 ~I0 24 I 8 T0~ 81 .900 .800 100 1.00 .050 1.00 1.00 105 .200 138 .050 1.00 1.00 1_L_ ~ ~ 1.00 T °-l N 1 10 1 10 ~24 I 8 T0~ 100 1.00 1.00 l.QQ l.QQ l.QQ l .QQ i.oo 1.00 T0 ~ 2 N 1 10 1 TO 24 1 6^ 8~ 35 1. 00 ----___ 48 ~ .300 1.00 1.00 1.00 100 1. 00 1.00 . 700 1.00 GEN PRO N 1 TO I 10 24 1 8^ T0~ 88 -_--__ i.oo 1.00 100 1. 00 1.00 l.QQ l.QQ l.QQ i.oo I1E N T 6 I 6 18 I 8 8~ 50 " .722 8 5 -----1.00 100 1.00 1.00 1.00 .278 -

PAGE 34

28 Table 6 . Continued. Enzyme Allele tr 1 D > H H a tr 1 w < H > n < i-3 '-a > H O O s 'Z a GPD N 1 10 1 10 24 1 8 10 67 .100 76 1.00 100 1 . 00 1.00 1.00 1 . 00 .900 170 1.00 1.00 PGI N 1 6 1 7 24 1 6 8 61 1 .00 95 1.00 100 1.00 1.00 1.00 . 146 1.00 400 — .854 1. 00 HEX-DH N 1 8 1 10 21 1 8 9 83 — . 056 100 1 . 00 — 121 1.00 1 . 00 1.00 .976 1. 00 1 . 00 .944 144 — — . 024 ADK N 1 2 1 7 24 1 4 4 100 1.00 1.00 1.00 1.00 1.00 1. 00 1 . 00 .375 111 — .125 HBD N 0 5 0 3 23 1 5 7 100 1 . 00 ] . 00 1.00 178 1.00 1. 00 1. 00 BDH N 1 8 1 10 21 1 6 5 100 1.00 1.00 1. 00 1. 00 1 . 00 1.00 1 . 00 1. 00 ACPH-1 N 1 10 1 10 24 1 8 10 100 1. 00 1.00 1.00 1.00 1.00 1.00 1, . 00 1.00 ACPH-2 N 1 10 1 10 24 1 8 10 95 — 1. 00 100 1. 00 1. 00 1.00 1.00 1. 00 1.00 1. . 00 LAP N 1 10 1 10 24 1 2 8 100 1.00 1. 00 1.00 1.00 1 . ,00 1.00 106 1. 00 1. 00 _ _ _ EST-1 N 1 10 1 10 23 1 8 10 84 . 200 . 500 — 90 . 800 . 500 _ _ _ 1 . 00 92 1. 00 1. 00 95 . 087 100 1. 00 115 1.00 124 . 196 129 . 717 EST-2 N 1 10 1 10 24 1 8 9 86 1.00 89 1.00 91 1.00 100 1. 00 1.00 107 1.00

PAGE 35

29 Table 6. Continued. tr 1 a < i-3 hi Enzyme Allele AR i — i 25 H a » AL H IT 1 MO OT < H > n tr 1 tr 1 z a ALD N 1 6 1 824 1 710 100 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00

PAGE 36

30 Table 7 • Allele frequencies of Gomphus plagiatus , G. spicatus , 2 populations of G. exilis , and 2 populations of G. lividus . Enzyme Allele M M H H X X < < cn 1 1 *t 1 — 1 H 1 i + I 1 + n O H NJ H to to LDH N 41 10 16 3 19 2 12 14 80 1 . 00 1 . 00 1. 00 1.00 1.00 1.00 1. 00 100 1. 00 HEX N 42 10 15 3 18 2 12 14 100 1.00 — _ _ _ _ _ 102 1. 00 1.00 1.00 1. 00 1. 00 1.00 1. 00 IDH-1 N 44 10 15 4 19 2 12 14 100 1. 00 1. 00 1.00 1. 00 1.00 1.00 1.00 109 1. 00 — IDH-2 N 44 10 15 4 19 2 12 14 73 — — — — .042 .036 100 1. 00 1. 00 1. 00 1. 00 1.00 1.00 . 958 .964 PGM N 43 9 15 5 20 2 14 16 88 1.00 _ _ _ _ 91 _ . 500 . 600 . 525 .250 _ . 031 100 1 . 00 . 500 . 400 .475 . 750 1.00 .969 ADH N 44 10 18 3 21 2 14 16 81 . 167 . 024 . 500 . 063 100 _ 1.00 .833 .976 . 500 1.00 .937 138 1.00 163 1.00 TO-1 N 41 10 16 3 19 2 12 14 100 1.00 1.00 1.00 1. 00 1.00 1.00 1.00 1. 00 TO2 N 39 9 13 3 16 2 10 12 35 . 039 48 . 961 1.00 1.00 1.00 1.00 1.00 1.00 1. 00 GEN PRO N 44 10 16 4 20 2 14 16 100 1. 00 1.00 1. 00 1.00 1 . 00 1.00 1.00 1.00 ME N 35 10 12 3 15 2 8 10 45 1.00 90 1.00 1.00 1. 00 1.00 1 . 00 1.00 133 1.00 GPD N 44 10 15 4 19 2 12 14 72 1.00 100 1.00 1.00 1.00 1. 00 1.00 1.00 1.00 PGI N 32 9 14 4 18 2 12 14 53 . 036 . 028 61 1. 00 100 1.00 . 964 1.00 . 972 . 750 1.00 . 964 425 . 250 . 036

PAGE 37

31 Table 7. Continued. Enzyme Allele M H < a w H H X X <^ < 01 1 I It* \t H > H i 1 + 1 1 + O o i— > to to to to HEX-DH N 45 10 15 5 20 2 14 16 100 1.00 . 933 1 . 00 . 950 1. 00 1.00 1.00 121 1.00 .067 .050 ADK N 18 7 9 3 12 1 6 7 100 . 972 1.00 1. 00 1. 00 1.00 1.00 1.00 1.00 143 .028 HBD N 20 8 13 4 17 2 9 11 74 . 077 .059 100 1.00 1.00 . 923 1. 00 . 941 1.00 1.00 1.00 GDH N 34 6 15 3 18 1 11 12 100 1. 00 1.00 1.00 1. 00 1.00 1.00 1.00 1.00 ACPH-1 N 43 10 17 4 21 2 14 16 100 1.00 1.00 1. 00 1. 00 1.00 1.00 1.00 1.00 ACPH-2 N 43 10 17 4 21 2 14 16 100 1.00 1.00 1. 00 1. 00 1.00 1.00 1.00 1.00 LAP N 26 9 15 4 19 2 12 14 100 1.00 1.00 106 1.00 1. 00 1. 00 1. 00 1.00 1. 00 EST-1 N 42 10 15 5 20 2 13 15 84 . 500 . 578 . 567 90 1.00 .933 L. 00 .950 .250 .269 . 267 92 .179 . 067 . 050 95 . 250 . 153 .166 98 . 821 EST-2 N 43 10 16 5 21 2 14 16 86 1.00 1.00 1.00 100 1. 00 1. 00 1.00 135 1. 00 171 1. 00 ALD N 46 9 17 5 22 1 16 17 100 1. 00 1.00 1.00 1 . 00 1. 00 1.00 1.00 1.00

PAGE 38

Table 8. Allele frequencies of Gomphus hodgesi, G. australis , G. br imleyi , G. diminutus , and 2 populations of G. cavillar is . Enzyme Allele n n > > < < a > til o a o CO H i H O > n < % + D H + H 2 H LDH N 4 1 31 21 10 31 62 9 80 1 . 00 1 . 00 1 .00 100 1.00 1.00 1.00 1.00 1.00 — HEX N 3 1 29 21 10 31 60 9 102 1 .00 1 . 00 1. 00 1. 00 1.00 1.00 1.00 1 .00 IDH-1 N 4 1 32 21 10 31 63 9 100 1 . 00 1 . 00 — 1 . 00 147 1. 00 1.00 1.00 1.00 1. 00 _ IDH-2 N 4 1 32 21 10 31 63 9 100 1 . 00 1 . 00 1. 00 1.00 1.00 1.00 1.00 1 . 00 PGM N 4 1 32 21 10 31 63 9 91 1 . 00 _ — _ _ 100 1 . 00 1.00 1. 00 1 . 00 1.00 1.00 1 .00 ADH N 4 1 32 21 10 31 63 9 100 1 . 00 1 . 00 1.00 . 976 1.00 .984 . 992 1 . 00 138 — _ . 024 .016 . 008 TO-1 N 4 1 3 3 21 10 31 64 9 100 1 . 00 1 . 00 1.00 1. 00 1.00 1 . 00 1.00 1 .00 TO2 N 4 1 33 21 10 31 64 9 48 1 . 00 — 100 — 1, .00 1.00 1.00 1.00 1. 00 1 . 00 1 . . 00 GEN PRO N 4 1 32 21 10 31 63 9 100 1. .00 1. , 00 1.00 1.00 1.00 1.00 1.00 1 , . 00 ME N 0 1 32 21 10 31 63 9 100 1. ,00 1. 00 1. 00 1.00 1.00 1.00 1. ,00 GPD N 4 1 32 21 10 31 63 9 100 1. , 00 1. 00 1. 00 1.00 . 950 . 984 . 992 1 . 00 155 . 050 . 016 . 008 PGI N 4 1 32 21 10 31 63 9 100 1. 00 1. 00 1.00 1.00 1. 00 1. 00 1. 00 1. 00 HEX-DH N 3 1 29 21 10 31 60 9 100 1. 00 1. 00 1.00 1. 00 1.00 1.00 1.00 1. 00 ADK N 3 1 23 9 10 19 42 9 100 1. 00 1. 00 1.00 1.00 1 . 00 1. 00 1.00 1. 00 HBD N 4 1 32 21 10 31 63 9 100 1. 00 1. 00 1.00 1.00 1.00 1. 00 1.00 1. 00 GDH N 1 1 29 21 10 31 60 9 100 1. 00 1. 00 1. 00 1.00 1.00 1. 00 1.00 1. 00 ACPH-1 N 4 1 32 21 10 31 63 9 100 1. 00 1. 00 1.00 1.00 1.00 1 . 00 1.00 1. 00

PAGE 39

33 Table 8. Continued. Enzyme Allele n < n n > > > < + < < > a O a to H # H a m H i 1 + H s o i-3 2 H to to s H M IN i 4 i i j z C X 1 u j l b J inn 1UU i n n 1 . U U l n n L . U U l n n ± . U U i n c\ 1 . U U i n n i n n i n n 1 . 0 V 1 . o u LAP N 4 1 26 21 10 31 57 5 106 1 00 1 0 0 i no X • U W i o n i no. EST-1 N 4 1 29 20 10 30 59 8 86 . 500 92 . 500 . 100 . 033 .017 95 1.00 1.00 1 . 00 . 900 .967 . 983 98 .875 102 . 125 EST-2 N 4 1 33 21 10 31 64 9 81 .250 93 . 050 . 016 . 008 100 .750 120 1.00 135 1.00 1.00 .950 .984 . 992 144 1.00 ALD N 4 1 33 21 10 31 64 9 100 1.00 1.00 1 . 00 1 . 00 1.00 1 . 00 1 . 00 1.00

PAGE 40

3 4 Table 9 . Allele frequencies of 2 populations of Gomphus minutus , Progomphus n. sp. , and P. obscurus . >xS Enzyme Allele 2 3 i — i H i — i Z o h1 U) to 1 i + H K) M • n LDH N 28 2 30 4 18 80 1.00 1 . 00 1. 00 120 1.00 1.00 HEX N 17 2 19 4 16 92 — 1.00 1.00 102 1.00 1.00 1. 00 IDH-1 N 28 2 30 4 18 86 — 1. 00 88 — — — 1.00 100 1.00 1.00 1.00 — IDH-2 N 28 2 30 4 18 73 . 018 — . 017 — 100 . 982 1 . 00 . 983 160 — — 1.00 1.00 TGM N 28 2 30 4 18 92 — — — . 056 100 1 . 00 1 . 00 1. 00 103 — — — 1 . 00 . 944 ADH N 28 2 30 4 17 6 8 — — — . 029 75 — — — .971 100 1 . 00 1.00 1 . 00 — 163 _ — — 1.00 TO-1 N 28 2 30 4 18 56 — 1 . 00 . 941 100 1.00 1.00 1 . 00 167 . 059 TO2 N 28 2 30 4 18 34 1 . 00 1 .00 48 1.00 1.00 1. 00 GEN PRO N 28 2 30 4 18 88 1 . 00 1.00 100 1.00 1.00 1.00 ME N 20 2 22 4 15 42 . 125 100 1.00 1.00 1.00 155 . 875 1. 00 GPD N 28 2 30 4 17 100 1 . 00 1.00 1.00 170 1.00 1.00

PAGE 41

35 Table 9. Continued. Enzyme Allele s H H H 'Z y. O hH U 1 i rn U J UJ i 1 + c/i t— 1 M O PGI N 28 2 30 4 16 56 1.00 1. 00 61 . 179 .167 100 .821 1. 00 .833 HEX-DH N 27 2 29 2 13 95 1. 00 1.00 100 1. 00 1. 00 1.00 ADK N 28 2 30 2 11 100 1.00 1.00 1. 00 1. 00 1 . 00 HBD N 27 2 29 4 17 78 1 . 00 1 . 00 100 1.00 1.00 1. 00 GDH N 16 2 18 2 14 100 1.00 1.00 1.00 108 1.00 1.00 ACPH-1 N 28 2 30 3 12 100 1.00 1.00 1.00 1.00 1.00 ACPH-2 N 28 2 30 2 12 100 1.00 1.00 1.00 1.00 1. 00 LAP N 28 0 28 4 11 100 — 1 . 00 1.00 106 1.00 1.00 EST-1 N 24 2 26 4 18 90 .333 . 500 . 346 95 .667 . 500 . 654 117 .750 . Ill 127 .250 . 194 137 .472 155 . 223 EST-2 N 28 2 30 4 18 107 1.00 1.00 1.00 260 1.00 460 1. 00 ALD N 28 2 30 4 18 100 1. 00 1.00 1.00 126 1 . 00 1.00

PAGE 42

3 6 100 origin ^ *x} 3 an tr m tr m (d w -a f a rt < T) hi(ihii dj <; tr & a " ' H& hj C O HX YJ H O O HfD d PHP -J H0 a d pp3 rt >.q hh n uq d a h m o cu n> < 3 p< n oi tr rt 3 H H h 0) CL PCJ PH CD O HPrt 0) Hl < Q, pCD Offltiftpj^HH-oi d rt 3 hpd CD rt hro d p. CO d M H . H tn C rt « 4 C Hi3 « M cj 2 • tn pco cn d tc oi hj co O rt £ 0) 01 01 0) o a d CO Figure l. Relative allele mobilities of acid phosphatase-1 in species of Gotnphus and Progomphus .

PAGE 43

9 5 100 orig in »d 'fl 3 p. f) D' BJ H (B W 'O *0 rt < -0 Hi Q, M Tj & < tT DJ 3 n rn E ^5 »"« C 01 rtOJ |D tn Ul Qi Mi 0) ft Q. Hp. » ™ . M ? W Mtn' " £ C "C B K C Hi 2 (II n 01 CD W i-j w o rt U) in o 3 c 3 en in tn Figure 2. Relative allele mobilities of acid phosphatase in species of Go mphus and Progomphus.

PAGE 44

Ill 143 origin ' f9 ^'IeTo, o Ef cj ? h ni "uPo >_"?+ < f o h. c_ h^o cu < trl^ ^ • • HHQ) Hj _ O HX ^ M O O Hn> £ HH(J) h d HM tT • 33
PAGE 45

3 9 100 105 130 138-^~ ongm ^P 3 an o pj S' m ft) w'c^ rt <>a thQi M-o pi < tr dj 3~ w W w tn o 3 C 3 01 0) Figure 4. Relative allele mobilities of alcohol dehydrogenase in species of Gomphus and Progomphus

PAGE 46

40 100 126. origxn_ " ' K 2 2 £ OHX -O H O O Hpi C HHf» '0 H^ & ? | ? || B | -S a P 3 s. E I £ E 2. S 8 M £ 3 . o xi c en Figure 5. Relative allele mobilities of aldolase in species of Go mphus and Progomphus .

PAGE 47

^ >x) 3 • • HO 3 C ry • rt (/) c o in 0) C ^3 t-i • c ' (u ii C O H-X < HW Qj < HH3 rt '13 HM ID ^< f J a hen W 13 ^ 'OHO Hrt £J fD rt 2 < T! i-h Dj MtJ pj rt Qj fi Hi &> rt pi 'O l"f O < 3 rt h Hi H h. en O D w hh tr • H. ro H< Oi H H(/I Hi O 01 tr H. U) re U < • HDi rt Relative allele mobilities of esterase-1 in species of Gomphus and Progomphus .

PAGE 48

93100107120124125' 135' 140' 144' 17l' 260' 460' origin t t)'r)3(inD'[ii5'H(i)(i)'ij'o r t<'o o a tr • tn o w C H3 M • C tn pj < H M PO H'XU HO O Hd) rt £ h H H. PPPt/1 W P, < PHQJ rt iX) p|— ' n <-Q P fD P. P0J p C U) rt fu W C rt W C p. Hi p, C PH p* n cu pr+ Qj hh QJ P o P rt p. hrc Hi a P 01 o 3 a a> P o < 3 P1 ^ PP M CD H< P, PPW Hi h o a tn tr tr P U) fD TJ <; . pPJ rt c tn Figure 7 Relative allele mobilities of esterase-2 species of Gomphus and Progoinp hus .

PAGE 49

origin tntr 3 ptu o m D m (u en t p u rt < H rnrF "o m <~eru~rr . . M-H-MCOP'XHhOOH-IDCH.H-pjtJH-hff. n „^ 3 < H-B) & < H-P-D) rt? PHU H3 ^ O H (5 D' 0 3CM.H-3(tiQhHn^C3HHO[U(I)<3H-<4I)l tT • r+ 3 HHti fD QjH-CU hM n> O HHrt (U H-^ G, (D tn CCMdHMCWrtliKloitiiaHitJrtftM-h'-BK. O in to rl-pk; hp-01 C (tD h-h-c (t rtH-ffl C n, hC C M HH10 d rt DOl^CPDBlh! cu ^ * to hin wdfD c/iHjcnort Figure 8 . Relative allele mobilities of general in species of Gomphus and Progomphus . protein

PAGE 50

•14 100108origin '^3 CL O tX 0> UPlt w >0 '0 r+ < •« Hi a, M'd P) < tr 0) 3 • • p-p-ojucoKX'OHOOKPi e hpju a ph rr n „ 2 f 3 < t;01 a < PCU s; M H hj m ^ hj O H (D tr 0 3 C H'H'3 (tifl h-ho^h: 3 HHO Pi » cn h en o rt G 01 05 if, o 3 c 0) US 05 0) 01 Figure 9. Relative allele mobilities of glutamate dehydrogenase in species of Gomphus, and Progomphus.

PAGE 51

4 5 100 origin S rn m £ £.2 P" W C W f+ D> fD in W Q, Hi V rr D. P" M " • 2 m w wcri) tn H in o rt w w (n o 3 c en 3 tn tn en Figure JO. Relative allele mobilities of a-qlycerophosphate dehydrogenase in species of Gomph us and Progoraphus .

PAGE 52

origin O D tr • w O en C I-! • [fl p H rt qj k < C H. H (/) H01 C O H ftu3 H PJ h1 HHHCD i — 1 o uq oi rt dj C rt w c U) O O rt f_ C 3 i— 1 ro HOJ C HHH H H MS OJ (C rt OJ I— ' I— ' O o I • fD Qj Hi OJ rt a H(D Hi t3 H. C/) o 05 OJ H. o <: a H-'< Cj"3" tr • tr h w ro 'd < . HOJ rt Figure 11. Relative allele mobilities of hexanol dehydrogenase in species of Gomphus and Progomphus.

PAGE 53

102 orig in • • PPPj H c O PX T3 H O O P0J (J p. p pi -o p. 4 & . ri ~ 3 3 ^ m ° J < H 0) ft £ H H hi H 3 Hi O hi ft) CT £ 3 £trL-3rl^ H MniQC3MM() ^tt<3p.
PAGE 54

48 100178O 3 D' • 71 • P3 c rt d m PPJ hj C O 3 < P0) D PP3 rt iQ 3 HH^ (!) c p id n h rt P) K (— 1 P d P Pptn PCO cn X '0 MOO PPPJ r+ .< Hnin d 3 PJ PM CD rt pi (I) (ji d rt 3 pd rt [/> d tn PM H O w p CD CO 0) d P' P" PJ 'Cj P J P H 3 p O H O PJ (D <; g ppft PJ p1 ^ a Hi (!) rt d. pd (D rt P(D II) H C Hi 3 tn p w O d cn p tr • cd cr < P to PCD T3 W < • PPJ rt C cn Figure 13. Relative allele mobilities of b-hydroxybutyrate acid dehydrogenase in species of Gomphus and Progomphus . ~

PAGE 55

130 1 4 7 origin • • hO 3 C t? 1 • rt O U) U) h • c hm h c O p 3 -4 pcn Dj < hpcu rt g PH3 rt v£) HH O i£) C 3 3 H H M (D D C H (t (U IB C rt ft) < M H01 C H. PPcn hcn cn XtjHOOH-Ofl pHM t— 1 H, I— ' 3 o tu m DJ rt C cn PPJ O ro cn 3 p' rt C cn pPrt Dj Hi PJ *9 DJ ft) f "0 H o < 3 p-*< D. PfD Hi 3 H cn O ro PH h rt) p< Dj Ppcn Hi H. O 3 Ifl DJ 2 tr • tr H 0) (D '13 <; . Pft) rt C tn Figure J 4. Relative al lele mobilities of dehydrogenase-1 in species of Progomphus . isoc i f ra te Gomphus and

PAGE 56

origin ^ ^ 3 an cr cu trnra w 'u tj • • HH£U h( C O PX t) H1 O 3 3 < H01 <; HHDJ rt o 3c; H-H-3 rtin H-MOi£)C W CCHftUMClllrtDin) nmicrtdi^Hp'Oi c rt 3 C tl C h. w p03 C rt *-t ' 01 H(I) 01 C C 03 o) 0) r» <; 'o hi a h'o (u <; tr pi p O PPJ HPfu 'U HHj cr • ?J H M Hj K' -J t-i O M Q tr 3P J P J OCJn)<3P- rt a PP03 < • PH-C (D rt P(!) C Hi H'O M M C Hi D 0) ^ 0) (D 03 i--; 03 o rt w O 3 c 3 03 0) 03 Figure 15. Relative allele mobilities of isoeitrate dehydrogenase-2 in species of Goniphus and Progomphus . '

PAGE 57

origin T) ^ 3 Ch o tr fu D-pft in i ti ti • • KP-Ill ^ c o H-XD HO D 3 < HID & < PP& it O 3 C PP3 rP '-Q P M O '£) D' • rt 3 H H K (t Q, pm hh nmiortt»'<:HP-tii c (p m CXJ C M PpWCr+ h ( • II) M1/1 f/) C n< T3 HiCil-"0 oj < tr 0> 3 o p(i) c phoj 'o hHj tr • 5] H H H H 3 P O P (t D' 3 H H O DJ O < 3 p< p w ro o pp(+ 0) p->< Qj pn> u (/) W a Hi [D rP O. PPtn < • PP(J 0) (t p(D C Hi p'OUjPCHiDUjH DJ fl) in ^ in o ft tn O 'J rj 3 U) in w Figure 16. Relative allele mobilities of lactate dehydr genase in species of Oomph us and Progom phus .

PAGE 58

222 origin ^ ^ 3 • • H3 O 3 C tr • rt W 3 O U) w i-l • g <: tr CD D'-ICO HW Qi 3 rt iQ H *1 (t CD ft) U) < H H01 M CD W '(J ^ M X MO < HHft) rt H-Mmfl 3 fj HM rt ft) CD 3 rt 3 3 rt tn c tn rt 0 3 (D < Xi t-h Hi M'rj ft) Hpi C Hhft) 'rj M M 'i M3 ^ O M M O ft) fD < g Hft ft) H-MJ Hi ft) rt Qi pro rt p cd c H 3 Hi y tn tn h, tn O i': O H Cn Q 3 tn ft) I ' h; H< Qj H M tO Hi l-{ 0 3 tr ft) 3 H D' • tr h tn • ro 13 <; . Vru rt 3 Figure 17. Relative allele mobilities of malic enzyme in species of Gomphus and P rogom phus .

PAGE 59

136 origin • • HHCJ H d O HX -0 H O O Hpi d HH£u >b H^ • n m 2^ ^ t !! a ^ P ' p ' ^ rt ^ p H ^ H a M O H O D' 03CH-H-3rttQH-HOiQC»MMOni{D<3H-
PAGE 60

54 400 origin •d g D,n tr DJ IT p (D to 'O rt < >a t-h & p"d ju • • P-H-ttHCOH-X-OHOOP-mCKH'P) '0 r, m 2 , 3 ^ 05 ^ < H(U r HH H ^ p U o 3 t hh3 rtiQ Hh n ui c d h h n ci it ^ y tr • rt3 HHtl ffi & V-P) HH (t O HPft C) P-^ w CdH(0[stlidOl(tP)(IifflfflD, HiWrttt POWU>rtJ>JK:HH-W C rt 3 PPc fD rt PfD C CO C Hj p. p. II) d rt *d W H C Hi 3 w 2 * 05 P05 tji £ (X> 05 Hj Ul C 05 05 05 05 O tn tr tu 3 HHi tr 1 1 fD tr p. 01 P-. P(D •o H 05 <, l-h pHi 0 rt c 01 0) Figure 19. Relative allele mobilities of phosphoglucose isomerase in species of Gomphus and Progomphus.

PAGE 61

origin [n M » S W g M § j? M9Ure oxiaailY"^ or tetrazoli™ "athoaal S,;^ ° £ 222*and Pro gorophus

PAGE 62

35 48. 100 origin TJ ^ 3~Dj n tr~(u ^ P it cn f o 'tTrt HHDJ M C O HX M O O O 3 tr • w n w m • c tr, 3 < . C H fD ft QJ C H 01 c/i Qi rt h ^ (0 D PJ W C M Ht/1 I1 H0) Hrt w rt <; fD '0 Hi Oj y-"0 W~^~ MOM < 3 HC Hi H H h M I — 0 H tn 0 tn n pi it HriDJ Hi fu rt(D ri hM C HfD l-h 3 M C/l O cn 3 tn "org-" tr •
PAGE 63

'0 3 [L O O' PJ Z J H (D in T3 '0 • • P • pPj hj C 0 PX H O 3 3 < Ptfl p.i < ppcu , t o y C PP3 rt l£! ppniQ c tX • (tSHHhdDiH-OH'H 0) CCHQdllJlCMrttDH O uj w rt P) ppen c rt 3 cm ppw c ,-f h w pen w C rt < *0 HiD, p"t) P) < CT P) S o ptu £ ppp -a pm tr • ? HHt1H3 II O n ID C 3P j Mncj(D "v m in £i pi pi ria. ppin < • ppC (C rt pfD c mi p. fD (/) p. cn O rt w O p c Relative allele mobilities of leucine aminopeptidase in species of Gomphus and Progoniph

PAGE 64

58 Table 10 . Percent loci polymorphic and percent loci variable for 22 alleles surveyed in Gomphus and Progomphus . % Polymorphic 5% Variable n. sp. abbreviatus brevis viridif rons apomyius parvidens lineatif rons dilatatus f urcif er pallidus villosipes townesi potulentus plagiatus spic atus exilis lividus hodgesi australis brimleyi cavillaris diminutus minutus P. n. sp. P. obscurus 76 63 10 6 5 1 9 1 9 23 1 7 9 39 9 19 14 4 1 31 30 9 28 4 16 9.1 4.5 9.1 13. 6 9.1 0.0 9.1 4.5 13.6 13.6 0. 0 0.0 18.2 4.5 0.0 18.2 9.1 9.1 0.0 0.0 0.0 4.5 9.1 9.1 13.6 22 . 7 18.2 13.6 13.6 9 0 9 4 13 18 0.0 0.0 18 . 2 13.6 0.0 27.3 22. 7 9.1 0.0 0.0 13.6 4 . 5 13.6 9.1 18.2 27.3 27.3 13.6 13.6 9.1 0.0 9.1 4.5 13. 6 18.2 0.0 0.0 18. 2 13 0 27 22 9 0.0 0.0 13.6 4.5 13. 6 9.1 18 . 2 Most common allele at a frequency of 95% or less. 3 Most common allele at a frequency of 99% or less. Percent of loci variable regardless of frequencies,

PAGE 65

si Q e o tn o u D, c rri U] si o, B 0 o 4-1 0 cn G) H +J iH W C Cn >i N O M 0) -P O XI (1) > T) C rj co o 0 ra U o H

PAGE 66

60 " sssSsssiissssiissiiilisiiiiisiiliissi 3 < i 1 I I I 1 I 1 gggggggg i i i i i i ° S ° ° o g 1 1 1 1 1 1 1 1 1 1 1 1 § § i § § § § i § 1 1 I a I I I I I I I I ooSSoggg s s s § g g \ § i § § i § § § i n 1 1 § § 1 1 1 1 1 1 1 £ a 1 I 1 1 5 I ; oooo 2934 0000 1245 0000 0950 6250 5711 5799 5000 oooo oooo oooo laoo 0644 0112 2199 4 4 4 4 5000 4527 1750 6775 < J 1 1 ! 1 1 I I I 55555555 I 1 1 ! 1 ! 1 5 s s s s s ; ! I I I I I I I 1 I I I I I I I I I I I I I I ° I < 1 I I I I I I I I I I I I I ! i 1 1 1 I 1 1 1 I III 1 1 II 1 1 1 1 1 11 I < I I I I I I I I I 1 I I I I ! : | I § § § I I § I § § I I | I § | § § | § I 1 § I I I I I I I I I I I I I I I I I I I I I I I 1 I I 1 I I 1 I | I 1 I I I I 1 I I I I I I I I ! I i 1 1 1 1 I I II I 3 I I I I I ! I I I I I I I I I I ° a I I I I I I I I I I I I I I I = 1 1 1 1 1 1 1 1 1 1 1 1 i 1 1 1 1 1 1 1 1 I s I I I I I I I I I 1 1 1 a i 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 o o IX .0000 . 1 JO J .0000 . 0718 . 1540 .1172 .0000 .0000 1 1 1 1 1 1 1 1 1 1 1 1 § 1 1 1 n i n i I § 1 I 1 I I 1 I 1 1 1 1 1 I 1 1 I 1 1 1 1 1 1 1 1 1 1 1 1 1 1 § 1 1 1 1 1 1 § 1 1 1 1 1 1 1 1 f I 1 I I s 1 o o o o * o 1 1 1 § § 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 = 1 I § 1 1 1 1 1 1 1 1 1 I ! 1 I 1 1 o o o o o o I I I I 1 I I I I I I ! I I I I I I 1 I ! I 1 s 1 1 1 1 1 1 1 1 ! I 1 ! 1 1 I o w o o o o I I I I I I I I 1 I I I 1 I 1 I 1 I 1 1 1 I z o i1 1 1 1 1 1 1 1 1 I I I 1 1 1 o o o o o o 1 1 1 1 I I I 1 1 1 1 1 1 1 1 1 1 1 i 1 1 1 s | < 1 1 I 1 1 § I I I s i s i I i O " o o o . 0000 .0000 . oooo .5000 .2778 .0465 . 5000 .0000 .1172 .0000 0000 0000 .0465 .0000 . .0)17 .0157 . .0000 .0000 .0000 . .0000 . .0000 .057 1 ° s lliililll S g g g g g S g 2 g S g 1 I II I ! 1 I I I I I I I I I I I I I I I o I I I I I I 2 1 I 1 I I I I I 1 i II 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 k 1 1 1 1 1 1 1 1 1 I I I I I I I 1 11 I I I 1 ! I I I I I 1 1 I I I I I I 1 I 1 1 1 1 1 1 § 1 f 1 I 1 I I I 1 1 11 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 I I 1 I I I 1 1 ! ! I I 1 I I 1 I 1 1 II 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 I \\ \l I I I I i 1 3 i 12 11 11 I | = Z I • ~ £ Z 11 .; | (X

PAGE 67

61 Table 12 Average heterozygosities, sample sizes, number of loci surveyed and number of variable loci in selected species of Odonata other than Gomphus and Progomphus . Species N No. No. V cl I X d UX fc: 1 1 n Epitheca cynosura 6 21 J . UDJ 5 E. Stella 6 21 2 o fi ? 8 E. costalis 5 21 E. sepia 2 17 o noon Libellula incesta 1 17 1 X n 9 Q A L. flavida 1 1 -1. . U j 1 z L. semif asciata 1 1 7 X / 1 L. auripennis 2 X o z . U 4 o 6 L. needhami 1 X 1 O 1 1 . 0278 0. ferruginea 1 L 1 7 X / n U .0000 Ladona deplanata 6 1 7 x / Till .1111 Plathemis lydia j 1 Q 4 . 0848 Pachydiplax longipennis 6 Didymops floridensis P-l 3 . 0 2 02 D. floridensis P-2 3 2 2 X n o n o * 0 20 2 D. transversa 10 .0127 Gomphaeschna antilope 1 1 8 X O n i 1 .0000 G. furcillafa P— 1 3 20 1 . 0139 G. furcillata P-2 11 22 1 . 0075 Epiaeschna heros 2 21 1 . 0233 Coryphaeschna ingens 3 IS 0 . 0000 Basiaeschna janata 2 20 0 . 0000 Anax junius 1 19 0 .0000

PAGE 68

Table 13. Matrix of unbiased genetic distance estimates for Gomphus and Progomphus based on analysis of 22 loci.

PAGE 69

63 1 2 3 4 1) a. SP. P-1 2) N. SP. P-2 . 0018 3) N. SP. P-3 .0167 . 0215 4) ABBR .1942 . 2034 . 1898 5) BREV . 2553 . 2580 .2484 . 2006 6) VIRI .0869 .0917 .0796 . 1422 7) APOM .1543 . 1603 .1468 .1033 3) PAW . 0444 .0494 . 0376 . 1006 9) LINE . 2549 . 2775 . 36 7 1 . 3624 LO) DILA . 1566 .1620 . 1669 .1860 1 I ) FURC . 8669 . 8637 . 7948 . 7978 1 2) PALL .6217 .6314 .6175 . 6222 13) VILL . 6866 . 6786 . 6324 . 78 30 L4) TOWN .6108 . 5777 .6555 . 6848 15) POTU .6012 . 5825 . 5416 . 7964 16) PLAG . 5916 . 5632 . 5734 .5903 17) SPIC . 3203 . 3114 . 3525 . 5265 18) EX P-1 . 2211 . 2254 .2450 .4010 1 ") EX P-2 . 2262 . 2281 . 2548 .4159 20) LIV P-1 . 3041 . 3025 . 3070 .4644 21) LIV P-2 .2865 . 2917 .2897 .4 390 22) HODG .2577 2658 . 2301 4007 23) AUST .2540 2593 . 2471 .3190 24) BRIM 2540 2593 .24 7 1 3190 25) CAV P-1 254 3 2591 . 2 4 74 3194 26) CAV P-2 2476 2532 .2 370 3133 27) DIMI 2487 2540 2418 3141 28) MIN P-1 2294 2328 2 346 3778 29) MIN P-2 164 5 1677 1752 3208 30) P N. SP. 1 683 1 675 1 678 1 681 31) P. OBSC 1 676 1 667 1 671 1 674 5 6 7 8 9 10 1443 1233 .0701 2735 . 0949 .1659 2387 .2387 .2366 . 3203 3146 .1232 .1201 .1869 . 0868 793 3 . 7962 .7133 . 8338 . 7207 . 7 3 32 6204 . 5168 . 54 32 . 5803 . 6137 . 4 7 3 9 6808 . 6839 . 6991 . 6 466 . 6890 . 6222 5938 . 68 39 . 5 27 3 .7419 . 6035 . 6683 6919 . 6004 .6172 .6516 . 598 3 .632 3 5061 . 5840 .4435 . 6354 . 5822 . 6 1 1 S 5136 . 3317 .4760 . 4055 . 4541 . 4940 4099 .2686 . 3524 . 2897 . 3311 . 227 1 40 36 . 2761 . 3674 . 3000 . 3249 . 2 4 3 3 4093 . 3022 . 4002 . 3421 .2289 . 2913 4 3 29 .2921 . 3805 . 3229 . 2696 . 2706 3146 .2513 . 3492 . 2365 . 3190 .1804 2618 .1499 .2331 .2719 . 2427 .1869 2618 .1499 . 2 3 3 1 .2719 . 2427 . 1869 2619 .1500 .2334 .2723 .2416 .1872 2615 1482 2310 .2657 2367 1805 3152 1881 2677 2663 2373 1813 274 1 2144 3019 2652 2805 2303 2275 1662 2509 4175 3719 3364 677 1 673 1 664 1 642 1 965 1 905 670 1 666 1 657 1 6 34 1 958 1 89!1 1 1 12) PALL . 4 304 13) VILL . 5566 14) TOWN . 8434 15) POTU .8533 16) PLAG .7212 1 7) SPIC . 6507 18) EX P-1 .8108 19) EX P-2 .8138 20) LIV P-1 . 7453 21) LIV P-2 . 8366 22) MODG . 7 320 23) AUST . 7320 24) BRIM . 3056 2 r >) CAV P-1 . 8045 26) CAV P-2 7914 27) DIMI 8002 28) MIN P-1 8313 29) MIN P-2 7413 30) P. N. SP. 1 660 31) P. OBSC 1 653 1 2 13 14 . 3486 .8693 .7835 8580 . 7917 . 390 3 7566 . 6894 . 2519 6528 . 5261 . 6931 624 3 . 6640 . 6770 6478 .6731 .67 31 6398 .6328 .70 35 6 397 . 6813 . 7503 4917 . 6222 . 6706 537 3 .6931 . 8938 6206 . 7885 . 7885 6216 . 7874 . 7850 6124 .7795 . 7845 5339 6378 .8885 6087 5092 . 7220 4 328 4640 .7312 4 39 1 466 1.466 4 32 1 459 1.459 15 16 17 3071 6896 .6894 674 1 . 6661 .2796 6787 . 6693 . 2737 6620 . 6291 .2739 6370 .6776 . 2887 4935 . 5635 . 2652 7962 . 7814 . 3830 6938 . 6859 . 4520 6955 . 6826 .4 509 6848 6799 4466 7909 7065 3 77 7 6990 6565 2917 5942 5803 2416 706 1 686 1 689 699 1 679 1 682 18 19 20 0000 0731 . 0688 0303 . 0869 . 0019 0301 .0316 .1065 2160 . 2241 .1829 3387 . 3475 . 3061 3392 . 3477 . 3058 3 326 .3418 . 3021 2107 2188 .1922 1351 1420 .1167 1257 1319 1172 954 1 957 1 917 947 1 950 1 909 22) 23) 24) 25) 26) 27) 28) 29) 10) 31) MODG AUST BRIM CAV P-1 CAV P-2 DIMI MIN P-1 MIN P-2 . N. SP. 21 . 1101 .1788 . 2958 . 2962 . 2910 . 1820 .1151 .1122 , 961 P. OBSC 1.954 22 . 2006 . 2652 . 2657 . 2567 .1813 .1216 .1270 1.911 1 .898 23 . 1466 .1468 . 1456 .1413 . 1566 .1106 1.977 1 .976 24 . 0000 . 0002 .2525 .2712 .2250 1.977 1. 970 25 . 0002 .1956 .2126 . 2253 1.976 1.969 2 6 27 2R 29 .1892 .2101 . 2218 . 968 1.961 .1334 . 0400 1.971 1. 964 30 . 0804 959 1.917 952 1.908 ,1683

PAGE 70

Table 14. Matrix of standard errors for genetic distance estimates of Gomphus and Progomphus .

PAGE 71

65 1 2 1) N. , SP. P-1 2) N, , SP. P-2 . 0017 3) N. SP. P-3 .0183 . 0213 4) ABBR . 0998 .1032 5) BREV .1173 . 1183 6) VI RI .0635 . 0654 7} APOM .0845 . 0863 8) LINE .1167 .1137 9) FURC . 2557 .2564 10) PALL .1981 .1999 11) VILL .2169 .2151 12) TOWN . 2002 .1903 13) POTU .1969 . 1910 14) PLAG .1952 .1862 15) SPIC .1342 .1312 16) EX P-1 . 1072 .1081 17) EX P-2 .1080 . 1089 18) LIV P-1 . 1295 .1300 19) LIV P-2 .1241 .1248 20) AUST .1169 .1182 21) BRIM .1169 .1182 22) CAV P-1 .1170 .1183 23) CAV P-2 .1145 .1159 24) DIMI .1151 . 1163 25) MIN P-1 .1083 .1088 26) P. N. SP. .4610 .4602 27) P. OBSC . 4603 .4595 10 11 11) VILL .1377 12) TOWN . 2582 .2390 13) POTU . 2527 .2397 14) PLAG .2335 .2175 15) SPIC .2076 . 1816 16) EX P-1 . 2006 .2110 17) EX P-2 .2068 .2147 18) LIV P-1 . 2078 . 2090 19) LIV P-2 . 2056 .2161 20) AUST . 1965 .2182 21) BRIM .1967 . 2390 22) CAV P-1 . 1951 . 2388 23) CAV P-2 . 1962 . 2374 24) DIMI .1799 .2172 25) MIN P-1 . 1960 .1775 26) P. N. SP. . 3982 .4009 27) P. OBSC . 3974 .4000 19 20 20) AUST . 0941 21) BRIM .1263 .0867 22) CAV P-1 .1264 .0868 23) CAV P-2 .1253 .0861 24) DIMI .0959 .0656 25) MIN P-1 .0737 .0706 26) P. N . SP. .5471 .5481 27) P. OBSC .5467 .5474 3 4 5 . 0986 .1153 .1026 .0599 .0842 . 0855 .0821 .0662 . 0757 , 1208 .1436 .1429 ,2398 . 2402 . 2395 ,1978 .1987 .1968 2163 . 2380 .2157 2075 . 2164 .1959 1844 . 2405 .2173 1910 .1956 .1762 1396 .1816 .1798 1109 .1512 .1537 1129 .1547 .1536 1313 .1685 .1534 1260 .1618 .1596 1148 .1336 .1198 1148 .1336 .1198 1149 .1338 .1199 1112 .1318 .1197 1132 .1322 .1326 1107 .1475 .1224 4609 .4608 .4604 4303 .4600 . 4597 12 13 14 .1512 .1161 .1291 .2182 .2175 .2175 .2138 .2121 .2114 .1225 .2147 .2159 .2139 .1199 .2226 .2146 .2087 .1223 .2286 .2172 .2158 .1242 .2390 .2194 .2167 .1650 .2378 .2182 .2156 .1647 .2384 .2149 .2154 .1635 .2379 .2170 .2155 .1642 .2614 .2396 .2212 .1475 .2192 .2192 .2085 .1243 .4009 .4637 .4615 .4616 .4000 .4630 .4609 .4609 21 22 23 24 0000 0004 . 0004 1006 .1008 . 0981 1051 .1052 .1048 5481 . 5480 . 5474 5474 . 5474 . 5468 6 7 0 9 . 0546 .1111 .1100 . 2404 .2210 . 2231 . 1722 .1804 .1957 .1578 .2165 .2192 .2158 .1841 .2165 .1815 .1982 .2491 .1981 .2016 .1967 . 2498 .1939 .1626 .1934 .2192 . 1486 .1683 .1651 .2063 .1197 .1379 .1329 . 2389 .1215 .1408 .1310 . 2411 .1301 .1515 .1117 . 2296 .1261 .1464 .1196 .2473 . 0880 .1091 .1124 .2417 . 0880 .1091 .1120 .2415 . 0881 . 1092 .1103 .2367 .0874 .1088 .1114 .2398 . 0976 .1178 .1107 . 2407 .1067 .1277 .1214 . 2462 .4602 .4598 . 5472 .4588 . 4596 .4592 . 5467 .4581 15 16 17 18 .0014 .0571 .0568 .0555 .0577 .0125 . 1049 .1073 . 0961 .1368 .1388 .1295 . 1369 .1390 .1296 .1348 .1370 .1289 . 1028 .1051 . 0999 .0769 . 0785 .0765 . 5462 . 5466 . 5447 .5457 .5461 . 5443 2 5 2 6 . 0800 .5477 .5468 .5472 .5463 .0921

PAGE 72

DISCUSSION Percent Polymorphism Percentages of loci polymorphic have been used as a measure of genetic variability in organisms studied electrophoretically. Of the various criteria that have been used to define polymorphic loci, the two most commonly employed have been to declare a system polymorphic if the frequency of the most common allele is equal to or less than 0.99 or 0.95. The percentage of all loci having more than one allele regardless of allele frequency is called the percent loci variable. Gorman and Kim (1976) found the percent loci variable averaged 35.5% for 15 populations of Anolis lizards. Three species of Bathygobius (inshore gobies) averaged 25.6% variable loci (Gorman et al. , 1976). Hubby and Lewontin (1966) found 43% of the loci variable in Drosophila pseudoobscura. The 23 populations of Gomphus averaged 11.3%. The highest percentage of variable loci (27.3%) was found in Gomphus ab breviates , G. n. sp. and G. exilis. By these comparisons Gomphus appear to be less variable than other organisms. However, Nei et al. (1975) point out that the percent loci variable, or polymorphic, is strongly dependent on sample size and therefore is not as good a measure of genetic variability as heterozygosity. 66

PAGE 73

6 7 Heterozygosity In random mating populations the average heterozygosity estimate is related to the proportion of an individual's loci that could be expected to be in a heterozygous condition. Even in nonrandom mating populations such as selffertilizing plants heterozygosity is a useful measure of genetic variability. In nonrandom mating populations the heterozygosity does not have a relationship to the frequency of heterozygotes . For this reason, Nei (1975) proposes substituting the term gene diversity for heterozygosity. As previously discussed, electrophoresis only detects about a third of the possible variation in the amino acid sequence of proteins. Therefore, the heterozygosity estimates presented are an underestimate of the total variability present in the genome. Selander (1975) summarizes heterozygosities from many studies. In Drosophila , 28 species had an H of 0.150, 4 other unspecified insect species averaged 0.151. Nine species of marine invertebrates averaged 0.147. Vertebrates tend to exhibit lower levels of heterozygosity. Twenty-six rodent species averaged 0.054. Webster et al., (1973) found 4 species of Anolis had a mean H of 0.0285, including one species with no variability. Berlocher (1976) found 16 species of Rhagoletis had a mean H of 0.1053. The 23 taxa of Gomphus averaged 0.0221, considerably below the heterozygosities of most other organisms that have been studied, especially other insects. Although the H for the 9 species

PAGE 74

68 of libellulids (0.0491) was twice the value for Gomphus it is still less than half as variable as the other invertebrates cited above. The low levels of variability in dragonflies may be attributable to their restricted habitat preferences, predatory nature, relatively small population sizes, and the possibility that they have undergone frequent genetic bottlenecks. Conversely, the migratory ability of most adult dragonflies would tend to offset, to some degree, the factors decreasing heterozygosity. The greater variability in the libellulids than in Gomphus may be correlated with a wider range of habitat preferences and generally larger population sizes. Nei (1975) presents the following equation for estimating effective population size from heterozygosity: H = 4N V e Where H = average heterozygosity N e = effective population size V = mutation rate. Kimura and Ohta (1971) suggest an average mutation rate of -7 10 per year for most proteins. Using this value and the H of G. abbreviatus , and solving for N , the effective population size is estimated to be 41,000. For G. pallidus the estimate is 129,250, for the average gomphine 54,500. The effective population size refers to the size of the panmictic breeding unit. It represents some quantity between the local deme and the total species population. The greater the gene flow between demes the closer the N will

PAGE 75

69 be to the total species population. From my personal field experience, these N g estimates seem far larger than the size of local populations, indicating gene flow is probably significant . Nei et al. (1977), and Chakraborty and Nei (1975) show quantitatively that the occurence of genetic bottlenecks can reduce levels of heterozygosity. The degree of reduction is related to the severity and frequency of bottlenecks, mutation rate and rate of population growth after bottleneck. The number of generations required for return to equilibrium is roughly the reciprocal of the mutation rate. Even with a mutation rate of only 10~ 6 that would mean a recovery period of a million years. Many Gomphus are obligatory stream dweLlers and presumably have been that way far back into their evolutionary history. The recent Pleistocene glaciations with their drastic effects on climate and sea level fluctuations probably had a disturbing effect on the restricted habitats of the Gomphidae. Considering the long period that would be needed to reach equilibrium, the low H observed today may be the sum effect of numerous glaciations and current small population size. If species could be sorted by susceptibility to the effects of glaciation future studies might be able to determine if in fact glaciation is responsible for the low levels of heterozygosity .

PAGE 76

70 Genetic Distance and Proposed Phylogeny Electrophoretic data can be used for both population genetics and systematics but the criteria required by the two disciplines can be quite different. Population genetics is concerned with the population dynamics within species, and large sample sizes are important to accurately estimate and follow small differences in gene frequencies. Systematists are interested in the evolutionary relationships between species and in the establishment and maintenance of a phylogenetically based classification system. Knowledge of the phylogeny within a related group can add considerable depth to the evaluation of comparative behavior studies. In population genetics it may sometimes be essential to confirm patterns of inheritance by breeding experiments. This is not practical with the majority of Odonata. The number of specimens required to establish the order of species within a phylogeny depends on the number of loci examined, level of heterozygosity and the genetic distance involved (Nei, 1978). The larger the distance between species the fewer specimens required. In organisms with high levels of heterozygosity large sample sizes would be necessary for construction of a dendrogram. Gomphus , with its very low heterozygosities should require minimum sample sizes. Gorman and Renzi (1978), working with Anolis roquet and Anolis bimaculatus species groups, demonstrated that the correct topology of a dendrogram could be achieved with even single individuals of each species.

PAGE 77

7 L The phylogeny proposed for all genera studied is presented in Figure 23. The general outline of the family groups within the dendrogram is complementary with the current taxonomic interpretation. All family groups clustered cleanly, without overlap. The Macromiidae (represented by Didymops) clustered with the five genera of Libellulidae at a distance of 2.111. These 2 families then clustered with the Corduliidae (represented by Epitheca ) at 2.576. Until recently, these 3 taxa had been considered only subfamilies of Libellulidae (Needham and Westfall, 1955). Gloyd (1959) elevated the Macromia group (including Didymops ) to family status. Corduliidae has also since been recognized as a family (Walker and Corbet, 1975, among others). The Gomphidae and Aeshnidae cluster at 2.859. Fraser (1957), in his reclassification of the order placed the Aeshnidae and Gomphidae together in the superfamily Aeshnoidea and the Libellulidae and Corduliidae (Fraser considered Macromiidae as part of the Corduliidae) in the superfamily Libelluloidea . This division is strongly supported by the clustering of the 2 superfamilies at 4.443. The ordering of the 3 families in the Libelluloidea may not be accurately depicted in the dendrogram. Gloyd (1959) states, with good evidence, that the corduliids and libellulids are more closely related to each other than to the macromiids. The relationship is just the reverse in the dendrogram. At such a large distance (2.1 2.5), a more extensive survey, using more genera, would be required for resolution of this question.

PAGE 78

T3 QJ in 4J rd c o o td 0 u m •H M -P QJ CD yi 0) o a 0 rH u 4-1 u o QJ H >1 0) c 1 X! 10 Qi OJ P rd f.l E 10 H O -P & [fl o Q) u a> a) u a) a si nJ -P +J cn •H 0 T3 6 U (d •H u +J tn 0) 0 c V-l CD tJ tn c c a o m
PAGE 79

73

PAGE 80

74 Gomphus and Progomphus cluster at 1.783. Progomphus is placed in the subfamily Gomphoidinae while Gomphus is in Gomphinae based on venational differences (Fraser, 1957) . The larvae of Progomphus and Gomphus are also very different. Therefore, the large genetic distance between them is not surprising . It is interesting to note that the 5 genera of aeshnids cluster at an average distance of 0.889, almost exactly the same as the average of the 5 genera of libellulids (0.893). This lends support to maintaining Plathemis and Ladona as valid genera rather than lumping them with Libellula as suggested by Ris (1910) and Kennedy (1922). Some current workers such as Walker and Corbet (1975) and Paulson and Garrison (1977) have adopted the usage of Ris and Kennedy. Needham and Westfall (1955) recognized both Plathemis and Ladona. Libellula depressa is the type species for the genus and is considered, on morphological grounds, to be most closely related to Plathemis lydia (Walker and Corbet, 1975). Any future attempt to resolve this question should include Libellula depressa . Figure 24 contains more details of the proposed phylogeny for the species of Corduliidae and Libellulidae surveyed. Orthemis ferruginea , placed by Fraser (1957) in the Libe-llul inae , clusters with Pachydiplax longipennis (subfamily Sympetrinae) before clustering with Libellula . The problems associated with subfamilies of Libellulidae and the resolution of species groups within Libellula could easily provide enough material

PAGE 81

0) 3 rH • rH fd CD -P X! rd -H U C H-> rd QJ 5-1 0) O •rH O •H U rH +J 3 O T3 QJ rH rH O i C tn qj CD tn +J O rd ^ e x: -p a w QJ (D tn o o c Cb rd O -P l-i 01 Cb-H cu x: u +j -H +j m a) o c QJ g Cn rd U C tn O O U T3 t3 oj C tn cu rd Q X! CN Q) rH tn H Dm

PAGE 82

76 V) < a to Z D < < z i/i 1 1— 1 UJ < 0 < « a UJ >hO Z D Ul u • to u u_ < z < J— < < u. < i q UJ >. ui _ f UJ < = < o z D < DC -* at a. Ul UJ u. a 6 J in z z UJ a, O z o

PAGE 83

77 for one or more additional dissertation research projects. I do not intend to draw any firm conclusions on the problems in Libellulidae. They were included only for comparison with the gomphines and for their obvious heuristic value. Many of the Libellulidae were represented by single specimens, a definitive study would require larger samples and the inclusion of more taxa, especially annectent forms. The 4 species of Epitheca clustered at a very low distance level (0.0461). The validity of these species was demonstrated by Tennessen (1973 and 1977) . The close relationship points out that species are defined in terms of reproductive isolation rather than genetic distance. However, electrophoretic analysis can prove quite helpful in delineation of new species (Ayala and Powell, 1972 and Berlocher , 1976) . The phylogeny proposed for Gomphus , based on the electrophoretic analysis of 22 loci, is presented in Figure 25. The subgeneric relationships are summarized in Figure 26. The branch points between groups of species in the dendrogram are obtained from the average of the separate pairwise distances of the species within the groups, therefore, the distance of the branch point may not match the distance between a particular pair of species. All described species and one undescribed species of the subgenus Hylogomphus were represented in this study. G. parvidens proved to be most closely related (D=0.0438) to the new species. G. viridifrons and G. apomyius formed

PAGE 84

o H i i n i UJ r—> o \ yy O u 1; 0) 03 (15 03 ft e +J 0 rd o TJ o o •H -P >i CD u 0 0 0 Ch h 0 >i u A -p ft u a; 0) as 03 O e G< 0 0 u M <4-l ft 03
PAGE 85

79 in z N « 2 O 3 in ^ a. a a. a. a ^ -^^^J^i^^^Z^S.a o.ao.>9*a>.2rf*— oooj_£5=>< 2 ^-ir;-0 * "6 'd d 00 "6

PAGE 86

u Cn H

PAGE 87

81

PAGE 88

82 a tight cluster with each other at 0.0701. G. abbreviatus clustered with the first 4 species at 0.1556 and G. brevis joined at 0.2148, completing the subgenus Hylogomphus . G. lineatifrons and G. dilatatus , the only representatives of the subgenus Gomphurus , clustered at 0.0868. This subgenus contains some of the largest and strongest flying Gomphus , their under representation in this study is partially due to the difficulty associated with their collection . The next 3 populations were virtually identical at all loci. That G. brimleyi might be only a synonym of G. cavillaris has been suspected (Westfall 1974). He indicated they were structurally inseparable and that the only difference was that G. brimleyi was more darkly colored. Based on both the striking enzymatic and structural equality G. brimleyi should definitely be considered synonymous with G. cavillaris . The 2 populations of G. exilis were identical at all loci, they were closely related to G. hodgesi (D=0.0309). Specimens of G. lividus from Florida, at the edge of their range, were very similar to specimens from central New York (D=0. 0019) . The clustering of the 2 populations of G. minutus with G. diminut us presents some conceptual problems. G. minutus from NW Florida were, according to the enzymatic analysis, more closely related to G. diminutus (also from NW Florida) than they were to G. minutus from Gainesville, Florida.

PAGE 89

83 G. minutus and G. diminutus can easily be told apart on the structure of the male abdominal appendages and other morphological characters. Of 3 possible explanations, introgression seems the least likely since intermediates between the 2 species have never been reported and also because G. diminutus has never been collected in proximity to the G. minutus population. The second possibility is that G. minutus may be a complex of sibling species, one of which is closely related to G. diminutus . Although this hypothesis seems unlikely and has no other supporting evidence, it cannot be disproven without further study. The third and most likely alternative is that the relationship is not phylogenetically accurate. The population of G. minutus from NW Florida is represented by only 2 specimens, the peninsular population by 23. The 2 G. minutus from P-l were both very teneral, collected and frozen within a few hours of emergence. All other specimens in the study were fully matured adults. The teneral specimens may not have had their entire complement of adult enzymes fully functional. One of the findings of Anderson et al. (1970) was that protein composition changed with transition from larva to adult. The distance between G. diminutus and the larger, mature population (P-2) from Gainesville is probably closer to the actual relationship between the 2 species. G. australis clusters with G. minutus and G. diminutus at 0.1362. The clustering of the subgenera is summarized in Figure 26. Species in the subgenus Gomphus did not cluster

PAGE 90

84 cleanly. The G. cavillaris populations (including brimleyi ) joined Gomphurus before clustering with the other Gomphus. The Gomphurus G. cavillaris group then clustered with the Hylogomphus at D=0.2418. This assemblage then clustered with most of the other species in the subgenus Gomphus at 0.2803. G. spicatus , a somewhat abberent form, is the last species of Gomphus to cluster (0.3757). Needham (1948) wrote that the nymph of spicatus is somewhat intermediate, lacking a lateral spine on abdominal segment 6, a fact that once lead him to erroneously place spicatus in Arigomphus . Of the 5 subgenera treated in Needham and Westfall (1955) , Stylurus and Arigomphus are clearly separated from the other 3. In Stylurus , G. townesi clusters with G. plagiatus at 0.2519, and they cluster with G. p otulentus at 0.3487. In Arigomphus , G. pallidus and villosipes cluster at 0.3486, and they cluster at 0.4935 with G. f urcif er . Nei (1975) summarizes estimates of genetic distance between taxa of different rank. The range of interpopulational distances (local races and subspecies) in a wide variety of vertebrates and invertebrates was 0.000-0.351. In Gomphus the range was 0.0000 0.0191 (G. minutus P-l and P-2 were excluded for the reasons discussed above) . The range for sibling species in Drosophila (Hubby and Throckmorton, 1968) was 0.18 1.54. In Gomphus , G. lineatifrons and G. dilatatus (D=0.0868), and G. parvidens and G. n. sp. (D=0.0438) can be considered sibling species. E. cynosura,

PAGE 91

85 E. stella and E. costalis are all sibling species. The largest distance between any 2 of them was 0.0461. Some non-sibling species pairs of Drosophila have distances as large as 2.54 (Hubby and Throckmorton, 1968). This is approximately the level at which families of dragonflies associate. The range for generic clustering for all dragonflies was 0.5023 1.783. Hence Gomphus seems to exhibit less genetic distance at all taxonomic levels. The question then arises as to what ranking should be assigned Arigomphus and Stylurus, the 2 groups clearly differentiated from the rest of Gomphus by the analysis of proteins. The branch point separating Arigomphus from the rest of Gomphus was 0.8083, similar to the levels attained by other genera in the order. The species of Arigomphus also seem to be a morphologically homogeneous group with similar habits and ecological requirements. I feel Arigomphus should be elevated to generic rank as originally suggested by Needham (1948) . The situation concerning Stylurus is not as easily resolved. Stylurus separated from the other species of Gomphus at 0.6612. This group is not, in my opinion, as morphologically or behaviorally distinct as Arigomphus . Walker (1958) also points out that the European G. flavipes and G. ubadschii may be intermediate between Stylurus and the rest of Gomphus. It is probably best to reserve judgement until future electrophoretic studies can include these species and more of the other North American species in Stylurus that were not represented in this study.

PAGE 92

86 Faster evolving enzyme systems with higher heterozygosities such as EST-1 tend to contribute more to the estimate of genetic distance than monomorphic systems. Distance criteria used to assess taxonomic rank are strongly dependent on the loci included in a study. Thus a generic level distance from one study is not likely to be equivalent to the same measure from another study using a different group of loci. The solution to this problem is the inclusion of a small number of standard reference species in all studies. One of the primary objectives of future studies in dragonflies should be to increase the resolution of enzymes with high heterozygosities, which would result in an increase in the accuracy of proposed phylogenies. Suggestions for future studies include determination of the generic limits in Libellula , further delineation of subgenera in Gomphus and clarification of the subfamilies of the Libellulidae. In the Zygoptera, studies of Enallagma and Argia would be of interest. Neutralists Versus Selectionists Considerable controversy revolves around the mechanisms responsible for maintenance of polymorphic enzymes in natural populations. Johnson (1974), Kojima and Tobari (1969) and others represent the selectionists. They feel most or all enzyme polymorphisms are maintained by some form of Darwinian selection such as heterosis or frequency dependent selection. The neutralists, lead by King and Jukes (1969), Kimura and

PAGE 93

87 Ohta (1974) and Nei (1975) feel that the majority of the observed variation is selectively neutral and the result of an equilibrium between mutation and genetic drift. This mutation equilibrium hypothesis does not preclude some small percentage of allozymes being maintained by selection, making it very difficult to disprove. Nei (1975) presents a convincing compilation of arguments favoring the mutation equilibrium hypothesis. One of the most interesting correlates of the mutation equilibrium theory is that a given protein will accumulate variation at a constant rate through evolutionary time, implying the existence of a rough protein clock. Aside from the problem of calibration of the molecular clock the question of neutrality versus selection has little effect on the main objective of this study, the computation of genetic distances and construction of phylogenies. Evolutionary Clocks Nei (1972) proposed that the genetic distance might be proportional to time and that it could be used as a rough, statistical clock for estimating time since divergence. For an ideal molecular clock, all proteins would have to be evolving at a constant rate. This is certainly not the case (Sarich, 1977,. Each protein is in itself a miniature timePiece. To achieve a reasonable estimate, a large number of loci are sampled with the hope that the different rates will average out. As much as possible the proteins to be analyzed

PAGE 94

83 should be a representative sample of the genome, selected without regard to evolutionary rates. In practice, in a preliminary study such as this, that usually means all systems that resolve well enough to be scored. Calibration settings of the clock have varied considerably. Nei (1975) sets 1 distance unit equal to 5 million years. Maxson and Wilson (1974) suggest a value of 18 million years estimated from an electrophoretic distanceimmunologic distance correlation. Gorman et al. (1976) hope to use the rise of the Panamanian isthmus to calibrate the clock. They are studying populations of fish from both the Atlantic and Pacific coasts that were apparently separated by the rise of the land bridge. The first group of species, on which they have reported, appear to be accumulating genetic distance at a rate consistent with the 18 million year calibration factor. Tillyard (1917) indicated the Gomphidae and Aeshnidae existed as distinct groups all the way back into the early Cretaceous, a span of approximately 130 million years. The genetic distance for the split of Gomphidae and Aeshnidae was 2.859, indicating that 1 distance unit equals 45.5 million years. This value seems very large and may be due to an underestimation of the distance between the families. At such large genetic distances (greater than 2-3) there is an increased probability that back mutations will reverse previously accumulated differences, resulting in distance values biased towards similarity.

PAGE 95

89 SUMMARY Protein variation in the dragonfly genus Gomphus was studied using starch gel electrophoresis. A phylogeny was proposed for 23 species using genetic distance estimates derived from analysis of 22 genetic loci. An additional 23 species of Odonata from a wide variety of families were analyzed for comparison with Gomphus. Average heterozygosities (H) for 23 species of Gomphus ranged from 0.0000 0.0852, with a grand mean of 0.0221. Six species of Gomphus had no apparent variability. Nine species in Libellulidae had a mean H of 0.0491. These H values are much lower than most other insects that have been investigated. The range of genetic distance levels for local populations was 0.0000 0.0191, for sibling species 0.0438 0.0868 and for genera 0.5023 1.783. In general the proposed phylogeny agreed rather well with the existing classification of Odonata as established from conventional morphological studies. On the basis of genetic distance Arigomphus was raised from subgeneric to generic rank. Gomphus brimleyi, identical at all 22 loci with G. cavillaris , was synonymized.

PAGE 96

90 Preliminary evidence based on small sample sizes indicates Ladona and Plathemis should not be lumped into the genus Libellula.

PAGE 97

APPENDIX A Locality Data Arigomphus furcifer New York, Tompkins Co., Ithaca, Conservation Pond at Junction of Rt . 366 and Upper Creek Rd., 14 June 1976. A. pallidus Florida, Alachua Co., Gainesville, 8, 9 and 19 April 1977. A. villosipes New York, Tompkins Co., Ithaca, Conservation Pond at Junction of Rt. 366 and Upper Creek Rd., 14 June 1976. Gomphus dilatatus Florida, Liberty Co., Sweetwater Creek at S-270, 10 May 1976. G. lineatifrons West Virginia, Hampshire Co., North River at Rt. 45, 6 and 7 June 1976. G. australis Florida, Wakulla Co., Piggot Pond near Medart, 2 6 March 1977. G. brimleyi Florida, Wakulla Co., Piggot Pond near Medart, 26 March 1977. G. cavillaris P-l Florida, Bradford Co., Sampson Lake, 27 March 1977. P-2 Florida, Alachua Co., Melrose, Santa Fe Lake, 2 April 1977. 91

PAGE 98

G. diminutus Florida, Santa Rosa Co., Big Coldwater Creek at S-191, 30 March 1977. G. exilis P-l New Hampshire, Grafton Co., Livermore, Church Pond 1 Km N of the Kancamangus Highway, 12 June 1976. P-2 New Jersey, Ocean Co., Friendship Branch at U.S. 70, 9 June 1976. G. hodgesi Florida, Santa Rosa Co., Big Coldwater Creek at S-191, 30 March 1977. plus: Florida, Santa Rosa Co., Pond Creek at S-191, 7 May 1976. G. lividus P-l Florida, Liberty Co., Crooked Creek at S-270, 6 May 1976. plus: Florida Liberty Co., Sweetwater Creek at S-270, 10 May 1976. P-2 New York, Tompkins Co., Cascadilla Creek at Turkey Hollow Rd. , 14 June 1976. G. minutus P-l Florida, Alachua Co., Gainesville, Newnans Lake, 25 March 1977, 1 and 8 April 1977. P-2 Florida, Taylor Co., Econfina River at Rt. 98, 2 6 March 1977. G. spicatus New Hampshire, Grafton Co., Livermore, Church Pond 1 Km N or the Kancamangus Highway, 12 June 1976. G. abbreviatus West Virginia, Hampshire Co., North River at Rt. 45, 6 and 7 June 1976. G. apomyius Mississippi, Forrest Co., Chaney Creek, 1 Km E of Brooklyn, 8 and 9 May 1976.

PAGE 99

93 2brevis Massachusetts, Middlesex Co., Squannacook River at Groton Conservation Area, 11 and 13 June 1976. G. parvidens North Carolina, Moore Co., Lakeview, Mill Creek, 2 June 1976. G. viridifrons West Virginia, Hampshire Co., North River at Rt. 45, 6 and 7 June 1976. G. n. sp. P-l Florida, Calhoun Co., Juniper Creek at Hwy 20, 6, 10 and 25 May 1976. P-2 Florida, Santa Rosa Co., Pond Creek at S-191, 7 and 25 May 1976. P-3 Florida, Liberty Co., Crooked Creek at S-270, 6 May 1976. plus: Florida, Liberty Co., Sweetwater Creek at S-270, 10 May 19 76. G. plagiatus Florida, Santa Rosa Co., Blackwater River at Bryant Bridge, 25 and 26 September 1976. G. potulentus Florida, Santa Rosa Co., Pond Creek at S-191, 25 May 1976. G. townesi Florida, Santa Rosa Co., Blackwater River at Bryant Bridge, 25 and 26 September 1976. Progomphus obscurus Florida, Calhoun Co., Juniper Creek at Hwy 20, 2 5 May 1976. P. n. sp. Florida, Calhoun Co., Juniper Creek at Hwy 20, 25 May 1976. Anax junius Florida, Putnam Co., West Lake, 0.5 Km N of Hwy 20, 14 March 1977.

PAGE 100

94 Basiaeschna janata Florida, Santa Rosa Co., Big Coldwater Creek at S-191, 30 March 1977. Coryphaeschna ingens Florida, Alachua Co., Melrose, Santa Fe Lake, 2 April 1977. Epiaeschna heros Florida, Alachua Co., Gainesville, Newnans Lake, 25 March 1977. Gomphaeschna antilope Florida, Liberty Co., Crooked Creek at S-270, 6 May 1976. G. furcillata P-l Florida, Wakulla Co., Piggot Pond near Medart, 2 6 March 1977. plus: Florida, Santa Rosa Co., Big Coldwater Creek at S-191, 30 March 1977. P-2 Florida, Alachua Co., Gainesville, Newnans Lake, 1 April 1977. Didymops floridensis P-l Florida, Putnam Co., West Lake, 0.5 Km N of Hwy 20, 14 March 1977. plus: Florida, Alachua Co., Melrose, Santa Fe Lake, 2 April 1977. P-2 Florida, Wakulla Co., Piggot Pond near Medart, 2 6 March 1977. D. transversa Florida, Wakulla Co., Piggot Pond near Medart, 2 6 March 1977. Epitheca costalis Florida, Wakulla Co., Piggot Pond near Medart, 26 March 1977. plus: Florida, Santa Rosa Co., Big Coldwater Creek at S-191, 30 March 1977.

PAGE 101

E. cynosura Florida, Wakulla Co. , Piggot Pond near Medart, 26 March 1977. plus: Florida, Wakulla Co., Lake Ellen, 26 March 1977. pins : Florida, Taylor Co., Econfina River at Rt. 98, 26 March 1977. E. sepia Florida, Liberty Co., Crooked Creek at S-270, 6 May 1976. plus: Mississippi, Forrest Co., Chaney Creek, 1 Km E of Brooklyn, 8 and 9 May 1976. E. Stella Florida, Alachua Co., Hawthorne, Johnson Lake, 14 March 1977. Ladona deplanata Florida, Santa Rosa Co., Big Coldwater Creek at S-191, 30 March 1977. Libellula auripennis Florida, Alachua Co., Gainesville, Newnans Lake, 19 April 1977. plus: Florida, Alachua Co., Melrose, Santa Fe Lake, 2 April 1977. L. f lavida Mississippi, Forrest Co., Chaney Creek, 1 Km E of Brooklyn, 8 and 9 May 1976. L. incesta Florida, Santa Rosa Co., Big Coldwater Creek at S-191, 30 March 1977. L. needhami Florida, Alachua Co., Gainesville, Stream near Bivens Arm on Rt. 441, 19 April 1977. L. semifasciata Florida, Alachua Co., Gainesville, 9 April 1977. Orthemis ferruginea Florida, Alachua Co., Gainesville, Stream near Bivens Arm on Rt. 441, 19 April 1977.

PAGE 102

96 Pachydiplax longipennis Florida, Alachua Co., Gainesville, Newnans Lake, 19 April 1977. Plathemis lydia Florida Wakulla Co., Wakulla River at Rt. 26 March 1977. plus: Florida, Alachua Co., Gainesville, Stream near Bivens Arm on Rt. 441, 19 April 1977.

PAGE 103

APPENDIX B Buffer and Stain Formulations Common chemical abbreviations used in the stain recipes are: ADP (adenosine 5' diphosphate), ATP (adenosine 5' triphosphate) , DH (dehydrogenase) , MTT (methyl tetrazolium) , NAD (nicotinamide adenine dinucleotide) , NADP (nicotinamide adenine dinucleotide phosphate) , NBT (nitro blue tetrazolium) and PMS (phenazine methosulf ate) . Distilled water was used in formulation of buffers and stains. All chemicals were obtained from Sigma Biochemical Company, St. Louis, Missouri. All recipes modified from Bush and Huettel (1972) or Brewer (1970) . Gel Buffers Histidine 8.0 Histidine HCL, 1.05 g in approximately 800 ml water, adjust pH to 8.0 with 2.0 N NaOH, dilute final volume to 1.0 liter. Poulik Tris, 152 ml of 1.0 M, 10 ml 1.0 M citric acid, in about 1.5 liters water, adjust pH to 8.6 with either 1.0 M Tris or 1.0 M citric acid, dilute to 2 . 0 liters. Dilute 1:1 when used in gels. Tris-HCl Trisma base, 1.21 g in approximately 900 ml water, pH to 8.5 with concentrated HC1 , dilute to 1.0 liter. 97

PAGE 104

98 Electrode Buffers Borate Boric acid, 36.4 g and 4.8 g NaOH in approximately 1.5 liters water, adjust pH to 8 . 2 with 1.0 M NaOH, dilute to 2.0 liters. Histidine 8.0 Sodium citrate, 58.82 g in approximately 900 ml water, adjust pH to 8.0 with 0.41 M citric acid, dilute to final volume of 1.0 liter. Tris Citrate Tris, 687 ml of 1.0 M, 157 ml of 1.0 M citric acid, and 1.0 liter water, adjust pH to 8.0 with either 1.0 M tris or 1.0 M citric acid, dilute to final volume of 2.0 liters . Stain Buffers Dehydrogenase Buffer (DH Buffer) Tris, 49.4 g and 12.5 ml concentrated HC1 in approximately 900 ml water, adjust pH to 8.4 with concentrated HC1 , dilute to stock volume of 1.0 liter. Dilute 1:3 just before use. Phosphate Buffer A Monobasic sodium phosphate, 5.52 g in 200 ml water, pH 4.6. Phosphate Buffer B Tribasic sodium phosphate (Na 2 HPC» 4 -7H 2 0) , 5.36 g in 100 ml water, pH 12.4. Tris Maleate Buffer A Trisma base, 24.2 g and 23.2 g Maleic acid in 1 . 0 liter water, pH 3.8. Tris Maleate Buffer B Sodium hydroxide, 80.0 g in 1.0 liter, pH 12.4. Stain Formulations Acid Phosphatase, ACPH-1 and ACPH-2 Polyvinyl pyrrolidone 100 mg, MnCl 2 100 mg, MgCl 2 100 mg , NaCl 300 mg, fast

PAGE 105

99 blue 40 mg, a-napthyl acid phosphate 260 mg, tris maleate buffer A 25 ml, tris maleate buffer B 12.75 ml, and 62.25 ml water. Adenylate Kinase, ADK Glucose 90 mg, NADP 25 mg, ADP 25 mg, MTT 6 mg, NBT 20 mg, MgCl 2 100 mg, hexokinase 160 units, glucose-6-phosphate dehydrogenase 80 units, and DH buffer 100 ml. Alcohol Dehydrogenase, ADH NAD 5 0 mg , NBT 10 mg , PMS 1 mg, 2-propanol 8 ml, and DH buffer 100 ml. Aldehyde Oxidase, AO Benzaldehyde 1 ml, NBT 20 mg , PMS 1 mg, and DH buffer 100 ml. Aldolase, ALP Glyceraldehyde-3-phosphate dehydrogenase 200 units, sodium f ructose-1 , 6-diphosphate 500 mg, sodium arsenate 156 mg, NAD 50 mg, NBT 30 mg, PMS 1 mg, and DH buffer 100 ml. Alkaline Phosphatase, APH Polyvinyl pyrrolidone 100 mg, MnCl 2 100 mg, MgCl 2 100 mg, NaCl 300 mg, fast blue 40 mg, a-napthyl acid phosphate 160 mg, fast blue 40 mg , and Poulik buffer 100 ml. Esterase, EST-1 and EST-2 Acetone 2 ml, 1-propanol 5 ml, fast garnet 80 mg, a-napthyl acetate 30 mg, b-napthyl acetate 30 mg, phosphate buffer A 40 ml, phosphate buffer B 40 ml, and water 20 ml. Fumerase, FUM Fumaric acid 200 mg, NAD 72 mg, NBT 20 mg, PMS 2 mg, malic dehydrogenase 200 units, and DH buffer 100 ml. Galactose Dehydrogenase, GAL-DH Galactose 1.8 g, NAD 50 mg, NBT 35 mg, NaCN 50 mg , PMS 2 mg , and DH buffer 100 ml.

PAGE 106

100 General Protein, GEN PRO Coomassie brilliant blue R, 100 mg in 100 ml fixative. Glucose-6-Phosphate dehydrogenase, G6PD Magnesium chloride 100 mg, glucose-6-phosphate 200 mg , NADP 70 mg , NBT 50 mg, PMS 1 mg , and DH buffer 100 ml. Glutamate Dehydrogenase, GDH Sodium glutamate 4 g, NAD 60 mg, NBT 30 mg , PMS 2 mg , and DH buffer 100 ml. Glutamate-Oxaloacetate Tranaminase, GOT Aspartic acid 440 mg, a-ketoglutar ic acid 240 mg , pyridoxal-5-phosphate 2 mg, DH buffer 100 ml, after ten minutes of incubation add fast blue BB 80 mg . Glyceraldehyde-3-Phosphate Dehydrogenase, G3PD Aldolase 200 units, sodium f ructose-1 , 6-diphosphate 546 mg , and DH buffer 20 ml, let stand 30 minutes then add: NAD 20 mg, MTT 20 mg, Na 3 As0 4 150 mg, PMS 4 mg, and DH buffer 100 ml. a-Glycerophosphate Dehydrog e nase, GPP a-Glycerophosphate 100 mg, NAD 12.5 mg, NBT 10 mg , PMS 1 mg , and DH buffer 100 ml. Hexanol Dehydrogenase, HEX DH Hexanol 5 ml, NAD 2 5 mg, NBT 2 0 mg, PMS 1 mg , and DH buffer 100 ml. Hexokinase, HEX Glucose 90 mg, ATP 25 mg , NADP 25 mg, MTT 20 mg, PMS 3 mg, MgCl 2 20 mg, and DH buffer 100 ml. b-Hydroxybutyrate Dehydrogenase, HBP Sodium chloride 800 mg, b-hydroxybutyr ic acid 500 mg, NAD 30 mg , NBT 20 mg, PMS 4 mg, MgCl 2 100 mg , and DH buffer 100 ml. Isocitrate Dehydrogenase, IDH-1 and IDH-2 Isocitric acid 200 mg, NADP 20 mg , NBT 10 mg , PMS 1 mg , MnCl 2 100 mg, and DH buffer 100 ml.

PAGE 107

101 Lactate Dehydrogenase, LDH Sodium D-lactate 2 ml of 1.0 M concentrate, NAD 50 mg, NBT 35 mg, PMS 2 mg , and DH buffer 100 ml. Leucine Amino Peptidase, LAP Black K 25 mg , L-leucyl b-napthyl amide HC1 20 mg, tris maleate buffer A 50 ml, tris maleate buffer B 10 ml, and water 40 ml. Leucine Amino Peptidase (Old) , old LAP Black K 25 mg , L-leucyl b-napthyl amide HC1 40 mg, phosphate buffer A 40 ml, and water 60 ml. Malate Dehydrogenase, xMDH Malic acid 50 mg , NAD 25 mg, NBT 20 mg, PMS 1 mg , and DH buffer 100 ml. Malic Enzyme, ME Malic acid 2 5 mg , NADP 10 mg, NBT 10 mg, PMS 1 mg, MgCl 2 100 mg , and DH buffer 100 ml. Monoamine Oxidase, MO Tryptamine HC1 1 g, sodium arsenate 156 mg, NAD 2 5 mg, NBT 2 0 mg, PMS 1 mg , glyceraldehyde3-phosphate dehydrogenase 200 units, and DH buffer 100 ml. Octanol Dehydrogenase, ODH NAD 2 5 mg, NBT 2 0 mg, PMS 1 mg, 1-octanol 0.2 ml, ethanol 1 ml 95%, and DH buffer 100 ml. Pepsinogen Soak gel for 10 minutes in bovine albumin 0.65%, pH 7.0, rinse with water and incubate 1 hour. Stain with normal general protein stain solution (above) . Phosphoglucomutase, PGM Magnesium chloride 200 mg , a-D-glucose-l-phosphate 170 mg , NADP 10 mg, MTT 10 mg, PMS 1 mg, glucuse-6-phosphate dehydrogenase 25 units, and DH buffer 100 ml. 6 Phosphogluconate Dehydrogenase, 6-PGD Magnesium chloride 20 mg , 6-phosphogluconic acid 20 mg, NADP 20 mg, NBT 30 mg, PMS 2 mg , and DH buffer 100 ml.

PAGE 108

102 Phosphoglucose Isomerase, PGI Magnesium chloride 100 mg, fructose6-phosphate 2 5 mg, NADP 15 mg , MTT 10 rag, PMS 1 mg, and DH buffer 100 ml. Sorbitol Dehydrogenase, SOR-DH D(-) sorbitol 500 mg, NAD 25 mg, NBT 20 mg , PMS 2 mg , and DH buffer 100 ml. Succinate Dehydrogenase, SDH Incubate gel slice in Na 2 succinate 13.5 g, K 2 HPO 25 ml 0.5 M , and 65 ml water. Pour off incubating solution and stain with Na 2 succinate 27 g, ATP 50 mg, NBT 35 mg, PMS 10 mg, NaEDTA 10 ml 0.1 M, and 0.5 M phosphate buffer (pH 7.0) 100 ml. Tetrazolium Oxidase, TO-1 and TO-2 NAD 25 mg , NBT 2 0 mg, PMS 5 mg , and DH buffer 100 ml. Expose to light until white bands appear on blue background. Triosephosphate Isomerase, TPI Sodium pyruvate 2.2 g, NAD 100 mg, a-glycerophosphate dehydrogenase 400 units, lactic dehydrogenase 400 units, and DH buffer 100 ml. Incubate gel in above solution for 2 hours, decant and adjust the pH to 2.0 with 1 N HC1 then back to pH 7.0 with 1 M tris. Stain with: NAD 60 mg, NBT 30 mg , Na 2 As0 4 250 mg, glyceraldehyde-3-phosphate 100 units, PMS 4 mg , and DH buffer 100 ml . Tyrosinase Cathechol 7.1 ml 0.01 M, L-proline 7 . 1 ml 0.01 M, and 100 ml 0.1 M phosphate buffer (pH 7.0). Xanthine Dehydrogenase, XDH Hypoxan thine 50 mg, NAD 20 mg, NBT 20 mg, MTT 20 mg, PMS 5 mg , and DH buffer 100 ml. Gel Fixative Methanol 5 parts, water 5 parts, and 1 part glacial acetic acid.

PAGE 109

LITERATURE CITED Anderson, M. , L. Halgren, and L. Nuti. 1970. Protein patterns of dragonfly hemolymph as shown by gel disc electrophoresis. J. Minn. Acad. Sci. 36:75-76. Avise, J. C. 1974. Systematic value of electrophoretic data. Syst. Zool . 23:465-481. Ayala, F. J. and J. R. Powell. 1972. Allozymes as diagnostic characters of sibling species of Drosophila . Proc. Nat. Acad. Sci. (USA) 69:1094-1096. Ayala, F. J. and M. L. Tracey. 1973. Enzyme variability in the Drosophila willistoni group. VII. Genetic differentiation and reproductive isolation between two subspecies. Heredity 64:120-124. Berlocher, S. H. 1976. The genetics of speciation in Rhagoletis (Diptera : Tephri tidae ) . Ph.D. Dissertation, Univ. of Texas at Austin. Brewer, G. J. 1970. An Introduction to Isozyme Techniques. Academic Press, New York. 186 pp. Bush, G. L. and R. N. Huettel . 1972. Starch gel electrophoresis of tephritid proteins. International Biological Programme. 55 pp. Chakraborty, R. and M. Nei. 1977. Bottleneck effects on average heterozygosity and genetic distance with the stepwise mutation model. Evolution 31:347-356. Fraser, F. C. 1957. A reclassification of the Order Odonata. Roy. Zool. Soc. N. S. W. Handbook, 12, 133 pp. Gloyd, L. K. 1959. Elevation of the Macromia group to family status (Odonata). Ent. News 70:197-205. Gorman, G. C. and Y. J. Kim. 1976. Anolis lizards of the Eastern Caribbean: A case study in evolution. II. Genetic relationships and genetic variation of the Bimaculatus group. Syst. Zool. 25:62-77. Gorman, G. C, Y. J. Kim, and R. Rubinoff. 1976. Genetic relationships of three species of Bathygobius from the Atlantic and Pacific sides of Panama. Copeia 1976361-364. 103

PAGE 110

104 Gorman, G. C. and J. Renzi, Jr. 1978. Genetic distance and heterozygosity estimates in electrophoretic studies: Effects of sample size. Evolution (in press). In M. Nei, Estimation of average heterozygosity and~genetic distance from a small number of individuals. Evolution (in press) . Harris, H. and D. A. Hopkinson. 1976. Handbook of Enzyme Electrophoresis in Human Genetics. North Holland Pub. Co., Amsterdam. Harrison, R. G. 1977. Parallel variation at an enzyme locus in sibling species of field crickets. Nature 266:168-70. Hubby, J. and R. C. Lewontin. 1966. A molecular approach to the study of genetic heterozygosity in natural populations. I. The number of alleles at different loci in Drosophila pseudobscura . Genetics 54:577-594. Hubby, J. and L. H. Throckmorton. 1968. Protein differences in Drosophila . IV. A study of sibling species. Amer. Nat. 102:193-205. Johnson, G. B. 1974. Enzyme polymorphism and metabolism. Science 194:28-37. Kennedy, C. H. 1922. The morphology of the penis in the genus Libellula . Ent. News 33:33-40. Kimura, M. 1969. The number of heterozygous nucleotide sites maintained in a finite population due to steady flux of mutations. Genetics 61:893-903. Kimura, M. and T. Ohta . 1971. Protein polymorphism as a phase of molecular evolution. Nature 229:467-469. Kimura, M. and T. Ohta. 1974. On some principles governing molecular evolution. Proc. Nat. Acad. Sci. (USA) 712848-2852. King, J. L. and T. H. Jukes. 1969. Non-Darwinian evolution. Science 164:788-798. King, M. C. and A. C. Wilson. 1975. Evolution at two levels in humans and chimpanzees. Science 188:107-116. Kojima, K. and Y. N. Tobari. 1969. The pattern of viability changes associated with genotype frequency at the alcohol dehydrogenase locus in a population of Drosophila melanogaster . Genetics 61:201-209.

PAGE 111

105 Maxon, L. R. and A. C. Wilson. 1974. Convergent morphological evolution detected by studying the proteins of the tree frogs of the Hyla eximia species group. Science 185:66-68. Needham, J. G. 1901. Aquatic insects in the Adirondacks. N. Y. State Mus. Bull. 47, Odonata, pp. 429-540. Needham, J. G. 1948. Studies on the North American species of the genus Gomphus . Trans. Amer. Ent. Soc . 73:307-339 Needham, J. G. 1951. Prodrome for a manual of the dragonflies of North America, with extended comments on wing venation systems. Trans. Amer. Ent. Soc. 77:21-62. Needham, J. G. and M. J. Westfall. 1955. A Manual of the Dragonflies of North America (Anisoptera) . Univ. of Calif. Press, Berkeley and Los Angeles. 615 pp. Nei, M. 1972. Genetic distance between populations. Amer. Nat. 106:283-292. Nei, M . 1975. Molecular Population, Genetics and Evolution. North Holland Publishing Co., Amsterdam. 288 pp. Nei, M. 1978. Estimation of average heterozygosity and genetic distance from a small number of individuals. Evolution (in press) . Nei, M., T. Maruyama, and R. Chakraborty. 1975. The bottleneck effect and genetic variability in populations. Evolution 29:1-10. Ohta, T. 1972. Population size and rate of evolution. J. Molec. Evol. 1:305-314. Paulson, D. R. and R. W. Garrison. 1977. A list and new distributional records of Pacific Coast Odonata. PanPacific Ent. 53:147-160. Prakash, S. and R. C. Lewontin. 1968. A molecular approach to the study of genetic heterozygosity in natural populations. III. Direct evidence of coadaptation in gene arrangements of Drosophila . Proc . Nat. Acad. Sci. (USA) 59:398-405. Ris, F. 1910. Collections zoologigues du Baron Edm. de Selys Longchamps. Fasc. XI, Libellul inen . Imprimeries des Academies, Brussels, pp. 245-384. Sarich, V. M . 1977. Rates, sample sizes, and the neutrality hypothesis for electrophoresis in evolutionary studies Nature 265:24-28.

PAGE 112

1U6 Selander, R. K. 1975. Genetic variation in natural populations. In F. J. Ayala, (ed.), Molecular Evolution. Sineaur Ass. Pub., Sunderland, Mass. p. 29-45. Selander, R. K. , M. H. Smith, S. Y. Yang, W. E. Johnson and T. B. Gentry. 1971. Biochemical polymorphisms and systematics in the genus Peromyscus I. Studies in Genetics VI. Univ. Texas Pub. 7103:49-99. Shaw, C. R. and A. L. Koen. 1968. Starch gel zone electrophoresis of enzymes. In I. Smith, (ed.), Chromatographic and Electrophoretic Techniques. Vol. II. W. Heineman, London. 524 pp. Stryer, L. 1975. Biochemistry. W. H. Freeman and Co., San Francisco. 877 pp. Tennessen, K. J. 1973. A preliminary report on the systematics of Tetragoneuria (Odonata : Cordul iidae) in the Southeastern United States. Unpubl . M. S. thesis, Univ. of Florida. Tennessen, K. J. 1977. Rediscovery of Epitheca costali s (Odonata:Corduliidae) . Ann. Ent. Soc . Amer. 70:267-273. Tillyard, R. J. 1917. The Biology of Dragonflies. Cambridge Univ. Press, London. 396 pp. Walker, E. M. 1958. The Odonata of Canada and Alaska. Vol. II. Univ. Toronto Press. 318 pp. Walker, E. M. and P. S. Corbet. 1975. The Odonata of Canada and Alaska. Vol. III. Univ. Toronto Press. 307 pp. Webster, T. P., R. K . Selander and S. Y. Yang. 1973. Genetic variability and similarity in the Anolis lizards of Bimini. Evolution 26:523-535. Westfall M. J. 1974. A critical study of Gomphus modestus Needham, 1942, with notes on related sp^cTis~7Anisoptera Gomphidae) . Odonatologica 3:63-73. Williamson, E. B. 1932. Two new species of Stylurus. Occ Pap. Mus. Zool., Univ. Michigan, 240. 18 pp.

PAGE 113

BIOGRAPHICAL SKETCH Kenneth William Knopf was born September 4, 1950, in Staten Island, New York. In 1967, after graduation from Port Richmond High School, he entered Cornell University, Ithaca, New York. In June, 1971, he received the degree of Bachelor of Science with a major in fisheries science. He was employed as a research biologist for Ichthyological Associates in Pottstown, Pennsylvania from June 1971 July 1972. In the fall of 1972 he enrolled in the Graduate School of the University of Florida. He has served as a research assistant in the Department of Entomology and Nematology for most of the past 5 years. In August 1974 he completed the degree of Master of Science. From that time until the present he has pursued work towards his doctorate. Kenneth William Knopf is married to the former Suanne Marie Saltsman. He is a member of the American Fisheries Society, the Florida Entomological Society, the International Odonatological Society, the North American Benthological Society and Alpha Gamma Rho . 107

PAGE 114

I certify that I have read this study and that in ray 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 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. Minter J. Iwestfall, Jr/ Professor of Zoology * 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. Lewis Berner Professor of Zoology This dissertation was submitted to the Graduate Faculty of the College of Agriculture and to the Graduate Council, and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. December 1977 Dean/[^ollege of Agriculture Dean, Graduate School


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EM5G9AA67_BWOJVU INGEST_TIME 2015-03-27T19:25:32Z PACKAGE AA00029787_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES