GENOME SIZE EVOLUTION IN THE INSECT ORDERS ODONATA AND
HYMENOPTERA
A Thesis
Presented to
The Faculty of Graduate Studies
of
The University of Guelph
by
ALEX ARDILA GARCIA
In partial fulfilment of requirements
for the degree of
Master of Science
August, 2008
© Alex Ardila Garcia, 2008 Library and Bibliotheque et 1*1 Archives Canada Archives Canada Published Heritage Direction du Branch Patrimoine de I'edition
395 Wellington Street 395, rue Wellington Ottawa ON K1A0N4 Ottawa ON K1A0N4 Canada Canada
Your file Votre reference ISBN: 978-0-494-47750-2 Our file Notre reference ISBN: 978-0-494-47750-2
NOTICE: AVIS: The author has granted a non L'auteur a accorde une licence non exclusive exclusive license allowing Library permettant a la Bibliotheque et Archives and Archives Canada to reproduce, Canada de reproduire, publier, archiver, publish, archive, preserve, conserve, sauvegarder, conserver, transmettre au public communicate to the public by par telecommunication ou par Plntemet, prefer, telecommunication or on the Internet, distribuer et vendre des theses partout dans loan, distribute and sell theses le monde, a des fins commerciales ou autres, worldwide, for commercial or non sur support microforme, papier, electronique commercial purposes, in microform, et/ou autres formats. paper, electronic and/or any other formats.
The author retains copyright L'auteur conserve la propriete du droit d'auteur ownership and moral rights in et des droits moraux qui protege cette these. this thesis. Neither the thesis Ni la these ni des extraits substantiels de nor substantial extracts from it celle-ci ne doivent etre imprimes ou autrement may be printed or otherwise reproduits sans son autorisation. reproduced without the author's permission.
In compliance with the Canadian Conformement a la loi canadienne Privacy Act some supporting sur la protection de la vie privee, forms may have been removed quelques formulaires secondaires from this thesis. ont ete enleves de cette these.
While these forms may be included Bien que ces formulaires in the document page count, aient inclus dans la pagination, their removal does not represent il n'y aura aucun contenu manquant. any loss of content from the thesis. Canada ABSTRACT
GENOME SIZE EVOLUTION IN THE INSECT ORDERS ODONATA AND
HYMENOPTERA
Alex Ardila Garcia Advisor: University of Guelph, 2008 Professor T. R. Gregory
This thesis provides genome size estimates for 187 species from two largely
overlooked insect orders: Odonata (dragonflies and damselflies) and Hymenoptera
(wasps, bees, and ants). Odonate and hymenopteran genome sizes ranged between 0.41 -
2.34 pg and 0.10-1.14 pg, respectively. In odonates, genome size did not correlate with body size, voltinism, or nymphal habitat. Interestingly, there was some evidence that
genome size correlates with flight ability; specifically, small genomes were associated with perchers and large genomes with fliers. In hymenopterans, parasitoid wasps contained some of the smallest genomes reported to date, but their genome size ranges were similar to that of non-parasitoid species. Surprisingly, there was evidence that
genome size did not correlate with social complexity. Genomes sizes in eusocial species were usually below 0.55 pg while solitary species displayed the entire range for the order. ACKNOWLEGEMENTS
I wish to express my deepest gratitude to my family and especially my parents in
Chile, Ricardo Ardila Macias and Fanny Garcia Rueda, for their moral and economic
support which was fundamental for the successful completion of my graduate and undergraduate studies at the University of Guelph. I am also most thankful to my
girlfriend, Polina Khrycheva, for all her support and advice during my studies at Guelph.
I am extremely grateful to my advisor, Dr. T. Ryan Gregory, for giving me the
opportunity to become his first graduate student and for all his guidance and time during the course of this project. I would like to extend my gratitude to all members of my
advisory committee: Dr. Brian Husband, Dr. Gary Umphrey, and Dr. Steve Marshall. I
am also most thankful to all members of the Gregory, Hanner, and Hebert labs, especially
Chandler Andrews, Paola Pierossi, Jillian Smith, Eugene Wong, and John James Wilson for their support and for making a great environment inside and outside university for the past two years.
Lastly, I would like to thank all the collaborators to this project. Paul Kron for his advice and guidance in flow cytometry: Colin Jones, Chris Early, Nancy Deyrup, Michael
May, Ed Lam, and Jerrell Daigle for their time and help in the collection and/or identification of odonates; Gary Umphrey, Jason Gibbs, Matthias Buck, Peter Kevan,
Mark Deyrup, and Stuart Fullerton for their time and help in the collection and/or identification of hymenopterans; Beneficial Insectary and BioBest Canada for their contribuition with access to parasitoid wasps.
i TABLE OF CONTENTS
LIST OF TABLES vii
LIST OF FIGURES ix
CHAPTER 1: An introduction to insect genome size diversity, methods, and
analysis 1
Abstract .2
Introduction.. 3
What is genome size, and why is it important? 3
Insects and their genomes 4
Objectives of the present study 6
Estimating genome size 8
Feulgen image analysis densitometry (FIA) 8
FIA and insect genomes '. 9
Challenges in FIA estimates in insects:. 12
Flow cytometry (FCM) 12
FCMin insects 13
Challenges in FCM estimates in insects 14
Summary 15 ii CHAPTER 2: Genome size diversity in relation to flight strategies in the insect order Odonata (dragonflies and damselflies) 26
Abstract 27
Introduction 28
Genome size diversity 29
Body size diversity.... 29
Developmental rate (voltinism) 30
Flight strategies ....32
Cytogenetics 35
Materials and methods 37
Collection, identification, and vouchering of specimens 37
Genome size estimation 37
Body size and wing area measurements .• 38
Voltinism and habitat 39
Karyotype data 40
Statistical Analysis 40
Results 40
Genome size diversity , 40
iii Body size and wing area relationships -. 41
Genome size vs body size 42
Odonate genomes vs nymphal habitat and voltinism 43
Odonate genomes vs karyotype 44
Odonate genomes vs flight strategies 44
Discussion 45
Genome size diversity.... 45
Body size 46
Nymphal habitat and voltinism 46
Cytogenetics 48
Flight strategies 49
Co-evolution of genome size and flight in odonates : 50
Genome size and flight 51
CHAPTER 3: Genome size diversity, parasitism, and eusociality in Hymenoptera: a first approach to the superfamilies Apoidea, Chalcidoidea, Ichneumonoidea, and
Vespoidea 89
Abstract 90
Introduction 91
iv Hymenoptera genomics 92
Genome size diversity 93
Objectives 94
Materials and methods 95
Collection, identification, and vouchering of specimens 95
Genome size estimation 96
Results 96
Genome size diversity 96
Genome size diversity among hymenopteran parasitoids 97
Genome size diversity at the family level in the superfamilies Vespoidea and
Apoidea ; 98
Discussion 100
Remarks on methods 100
Hymenoptera genome size diversity 101
Genome size in parasitoids 102
Genome size vs eusociality 102
Future studies 103
CONCLUSION 125 v REFERENCES..... 127
vi LIST OF TABLES
CHAPTER 1
Table 1.1 Insect genome size diversity 16
Table 1.2 General procedure in the Feulgen reaction for genome size estimation using
FIA 17
Table 1.3 FIA mean Integrated Optical Density (IOD) for internal and insect
standards 18
CHAPTER 2
Table 2.1 Odonata higher taxonomic classification including the number of genera and
named species found in North America 54
Table 2.2 Published odonate genome sizes 55
Table 2.3 Hypothesized relationships between genome size and body size,
developmental rate, flight strategy, and karyotype in Odonata 57
Table 2.4 Classification of Odonata according to developmental rate (voltinism) 58
Table 2.5 Classification of Odonata according to flight strategies 59
Table 2.6 Comparison between genome size estimates obtained from haemocytes and
sperm for six individuals from four odonate species 60
Table 2.7 Odonate genome size estimates for 62 dragonfly species and 38 damselfly
species 61
Table 2.8 Odonate body size measurements 71
vii CHAPTER 3
Table 3.1 Published hymenopteran genome sizes 105
Table 3.2 Genome size estimates for 87 hymenopteran species 107
Table 3.3 Genome size ranges and means for solitary and eusocial closely related
Apoidea and Vespoidea genera 118
Table 3.4 Genome size estimates for six ant species obtained in this study and
compared to estimates reported by Tsutsui et al. (2008) 120
APPENDICES
Appendix 1. Body size measurements with photographic image analysis 141
Appendix 2. Pearson correlations for odonate genome size and body size means 146
viii LIST OF FIGURES
CHAPTER 1
Figure 1.1 Diversity of named species of animal groups and distribution of genome
size estimates in animals 19
Figure 1.2 Number and percentage of publications reporting insect genome size
estimates since the first published estimate in 1952 20
Figure 1.3 Genome size diversity in insects 21
Figure 1.4 Genome size estimation using FIAD 22
Figure 1.5 Genome size estimation using FCM 23
Figure 1.6 Comparison of FCM measurements between the whole head and the
isolated brain ofAcheta domesticus (house cricket) 25
CHAPTER 2
Figure 2.1 Body size and wing area measurements of a representative odonate 86
Figure 2.2 Genome size and flight strategies in Odonata 87
Figure 2.3 Evolution of genome size, flight strategy, and karyotypes in dragonflies...88
CHAPTER 3
Figure 3.1 Phylogenetic relationships of the Apocrita superfamilies .....121
Figure 3.2 The phytogeny of "Aculeata" 122
Figure 3.3 Genome size ranges and means for hymenopteran superfamilies 123
ix Figure 3.4 Genome size ranges and number of species measured (n) for the Vespoidea
and Apoidea families sampled in this study 124
x Chapter 1: An introduction to insect genome size diversity, methods, and analysis
1 Abstract
This chapter gives an introduction to the long-standing puzzle of genome size diversity by reviewing current knowledge about genome size variation in insects. It provides detailed procedures for the estimation of insect genome sizes with the two major techniques used today— Feulgen image analysis densitometry and flow cytometry. Both methods were used in the present study to provide new estimates for 187 species from the orders Odonata (dragonflies and damselflies) and Hymenoptera (wasps, bees, and ants).
2 Introduction
What is genome size, and why is it important?
Genome size (GS), is defined as the total amount of DNA in a monoploid (lCx,
Greilhuber et al. 2005) chromosome set of a given organism. Genome size is typically
measured in mass (in picograms, pg) or number of base pairs (bp) of DNA, whereby lpg
= 10"12g = 978 Mbp (Dolezel et al. 2003). It has been recognized since the late 1940s that
genome size (though not necessarily genome copy number) is generally constant in all
cell types of an individual, and in all individuals within a given species (Gregory 2005).
In contrast, there is tremendous variation in genome size among eukaryote species:
animal genome sizes alone range more than 7,000-fold (Gregory 2008).
The large variation in genome sizes has been a major puzzle in science for nearly
60 years. Early on, it was considered surprising, even paradoxical, that complex
organisms such as humans do not possess the largest genomes. As just one of a large
number of possible examples, the genome size of the meadow grasshopper Chorthippus parallelus (GS = 13.36 pg) is nearly four times larger than humans (GS = 3.5 pg)
(Dole el et al. 2003; Wilmore and Brown 1975). This apparent paradox was solved with
the discovery and description of various types of non-coding DNA (sometimes called
"junk DNA") (Cavalier-Smith 1978), which meant that a large genome is not the result of
a greater number of genes, and thus need not be related to complexity in any obvious
way. However, the fact that a genome like that of humans consists of 98.5% non-coding
DNA raises several important questions that are still being debated today. Genome size
diversity is still considered a complex enigma, subdivided into several smaller questions: >
3 1) What are the sources of non-coding DNA? 2) How is non-coding DNA gained and lost
in genomes over time? 3) Does non-coding DNA have any effects or functions? 4) Why
are some genomes streamlined while others are large? (Gregory 2001, 2005). In fact, the journal Science recently dedicated a special issue (vol. 309 2005) to the 125 most
important questions in science which included two key questions in genome size
research: "Why are some genomes really big and others quite compact?" and "What is
all that "junk" doing in our genomes?"
Today, it is unclear what "junk" DNA is doing, but what is clear is that the size of the genome is strongly linked to cellular and organism-level phenotypes. At the cellular
level, genome size correlates positively with cell size and negatively with cell division rate (Bennett 1971, 1972; Cavalier-Smith 1978, 1982; Olmo 1983; Gregory 2001, 2005).
These cell level effects may result in several possible consequences at the organism level,
depending on the group in question. For example, genome size may correlate positively with body size in some groups (Finston et al. 1995; Gregory et al. 2000) or negatively with metabolic rate (Szarski 1970, 1983; Tiersch and Wachtel 1991; Hughes and Hughes
1995; Gregory 2002a) or developmental rate (Gregory et al. 2003; Gregory and Johnston
2008). These relationships show that genome size variation is not random and that its
characterisation is an essential step for the understanding of the evolution of animals.
Insects and their genomes
Despite the abundance and diversity of insects, insect genomes represent only
~10% of all available genome size data in animals (Fig. 1.1). This lack of information is
striking considering that insects represent -2/3 (~ 1 million species) of all eukaryotes
4 described to date and several million more species of insects probably await description
(Grimaldi and Engel 2005).
The first genome size estimate for an insect was that of Drosophila melanogaster
(GS = 0.17 pg), which was determined in 1952 by Kurnick and Herskowitz. Their results
were almost identical to that obtained by Rasch et al. (1971) (GS = 0.18 pg) 20 years
later, and further confirmed by whole-genome sequencing another 30 years later (Adams
et al. 2000). It took almost 20 years after Kurnick and Herskowitz's (1952) study before major research started on genome size variation in insects (Fig. 1.2). According to the
Animal Genome Size Database (Gregory 2008), there have been 531 published estimates
from 96 studies covering 471 insect species to date. Nearly 70% of the available insect
genome size estimates have been published over the past two decades, and more than 100 new estimates have been published in six reports over the two year duration of this study.
However, despite the growing interest and technological advances in the field, there are no large-scale efforts to characterize insect genome sizes.
Nevertheless, the limited data available have revealed that there is at least a 150-
fold variation in genome size among insects, ranging from about 0.11 pg in the twisted- wing parasite Caenocholax fenyesi texensis (order Strepsiptera) (Johnston et al. 2004) to
16.93 pg in the mountain grasshopper Podisma pedestris (order Orthoptera) (Westerman
et al. 1987). Most insect genomes have been found to be smaller than 1 pg (Fig. 1.3), but this must be examined cautiously because 75% of the estimates belong to just three insect
orders: Coleoptera, Diptera, and Lepidoptera (Table 1.1), which may be constrained in how much their genomes can vary in size (Gregory 2005; see following).
5 Thus, although current data do not provide conclusive information regarding the
genome sizes of most insect groups, they are sufficient to highlight extensive diversity
and suggest possible links to biological traits. Genome size has been reported to correlate
positively with body size in aphids (Hemiptera) (Finston et al. 1995) and negatively with
developmental time in mosquitoes, vinegar flies, and lady beetles (Ferrari and Rai 1989;
Gregory et al. 2003; Gregory and Johnston 2008). The most striking pattern in insects,
however, is the negative correlation observed between genome size and developmental
complexity across orders (Gregory 2002b, 2005). Specifically, holometabolous insects
(i.e., those that undergo complete metamorphosis) exhibit some of the smallest genomes
and appear to show a 2pg upper limit, whereas hemimetabolous (incomplete
metamorphosis) and ametabolous (no metamorphosis) orders tend to have larger genomes
and are not limited by any such threshold (Gregory 2002b, 2005).
Objectives of the present study
The shortage of genome size data in insects greatly limits our understanding of
genome size evolution. Few studies provide large scale surveys and most efforts have
focused on characterizing a few species, often those with economic or medical interest,
thereby limiting our ability to identify genome size variation patterns and examine their relationships to other traits.
A key aim of this study is to provide a large scale genome size survey of two
largely uncharacterized insect orders: Odonata (dragonflies and damselflies, Chapter 2)
and Hymenoptera (wasps, bees, and ants, Chapter 3). More importantly, the primary
objective is to use the new data to test specific hypotheses that are central to the study of
6 genome size evolution and that relate to some of the main characteristics of each of these
groups. It is important to clarify here that all hypotheses tested in this study were
generated based on previous observations that established a strong positive correlation
between genome size and cell size, and a strong negative correlation between genome
size and cell division rate (Bennett 1971, 1972; Cavalier-Smith 1978, 1982; Olmo 1983;
Gregory 2001, 2005).
Based on this, it was hypothesized that in Odonata genome size would correlate
positively with developmental rate and body size. In addition to testing these hypotheses,
the unique flying characteristics and strategies of Odonata allowed testing for the first
time in any invertebrate group whether genome size and flying activity correlate negatively as reported previously in vertebrates (Szarski 1970, 1983; Tiersch and
Wachtel 1991; Hughes and Hughes 1995; Gregory 2002a). This relationship is thought to be the result of the interaction between genome size, cell size, and metabolic rate; where
active fliers (high metabolic demand) evolved small cell sizes to maximize oxygen
uptake, and the reduction of cell sizes has had a similar effect on the size of the genome.
In Hymenoptera, the wide array of life strategies and social interactions that have
evolved in this group (parasitoid, solitary, primitively social, and eusocial) provided an
exceptional opportunity to examine for the first time two of the most intriguing
hypothesis in genome size research. Firstly, whether genome size is highly reduced in parasitoid genomes due to high developmental rate constraints; and secondly, whether
genome size increases along with social complexity. In this case, the relationship may be
due to restrictions to neuron size imposed by having more complex brains (more neurons)
in a limited space (Roth et al. 1994)
7 Testing of these hypotheses in odonates and hymenopterans in the following two chapters significantly expands our understanding of the evolution of the genome in insects, and animals as a whole. The methods used to obtain the 187 estimates generated in this study are outlined in the remaining sections of this chapter.
Estimating genome size
In principle, it would be ideal to sequence every nucleotide of a genome to determine its exact size. However, in practice, the presence of large amounts of non- coding DNA, especially repetitive sequences, makes this approach expensive, time consuming, and potentially inaccurate (Bennett et al. 2003). The most efficient, reliable, and cost effective methods for estimating genome size remain those that use measures of the relative content of non-base pair-specific stains or fluorochromes compared between unknowns and an established standard. The main methods currently used in genome size analysis, Feulgen image analysis densitometry (FIA) and flow cytometry (FCM), were both used in the present study.
Feulsen image analysis densitometry
Feulgen image analysis densitometry (FIA) represents the modern incarnation of the classical Feulgen densitometry method first used to assess genome size in the 1940s
(Vilhar et al. 2001; Hardie et al. 2002). It is based on the histochemical principle that
Schiff reagent binds proportionally to the amount of DNA found in the nucleus. Because
DNA is not distributed homogeneously in the nucleus, the amount of bound stain is determined for individual point densities and summed as integrated optical densities
(IODs). Until recently, Feulgen densitometry was a laborious process in which nuclei
8 were analyzed individually one point density at a time. The advent of FIA improved this
process considerably by introducing computer imaging to measure all point densities (in
every pixel) simultaneously for each nucleus found in a field and to calculate IGD for
each nucleus automatically using specialized imaging software (Fig. 1.4). This allows
dozens of nuclei to be measured together instantaneously.
FIA and insect senomes
Roughly half of the available insect genome size estimates have been obtained
using traditional Feulgen densitometry (Gregory 2008), but already 93 estimates have
been published using modern image analysis, with many more awaiting publication (T.R.
Gregory, pers. comm.; present study). The determination of DNA content using FIA
requires that the cell types selected are of known ploidy (number of genome copies per
nucleus) and can be easily dissected to obtain a single layer of cells fixed on a glass slide.
Tissues with these characteristics include spermatozoa (haploid) and haemocytes
(diploid), for example. Most specimens can be dissected fairly quickly, roughly 10
minutes per dissection, in the laboratory at a field station by using a pair of pins, a
dissecting microscope, a frosted glass slide, and a few drops of Ringer's saline (1L dH^O,
7.5g NaCl, 0.35g CaCl2,0.21g KCL).
In addition to preparing samples from insects with unknown genome content,
internal standards must also be prepared and co-stained in the same staining reaction. The best insect standards are those that can be accessed and maintained easily in the lab, and more importantly, those in which genome size has been established using multiple methods. Also, it is best to make estimates on the same tissue type in the standard and the
9 unknown. Today, the vinegar fly Drosophila melanogaster (GS = 0.18pg; Rasch 1971)
remains one of the most suitable standards for FIA when using spermatozoa.
Unfortunately, spermatozoa samples cannot always be obtained from unknowns,
meaning that haemocytes must be used, though these may be difficult to obtain from
small flies and hence other standards may be more suitable in some cases. The yellow
mealworm beetle (Tenebrio molitor; GS = 0.52pg; Juan and Petitpierre 1989) has been
used in previous studies, and is easily available and simple to rear in a lab. Finally, best practice methods involve the inclusion of internal standards in each staining run to
confirm stoichiometry, and for this purpose erythrocytes from Gallus domesticus
(chicken, GS = 1.25 pg; Rasch et al. 1971, Hillier et al. 2004) and Oncorhynchus mykiss
(rainbow trout, GS = 2.6 pg; Rasch 1985; Dolezel et al. 2003) are often used to confirm
IOD ratios.
Once prepared, samples are stained with Schiff reagent in the Feulgen reaction
(Hardie et al. 2002; Table 1.2). First, Schiff reagent (1000 mL dH20, 11 g of basic
fuchsin, 10 g of sodium metabisulfite, 100 ml 1.0 N HCL) is prepared 36 hrs before the
Feulgen reaction and kept at room temperature in the dark until use. Preparing 1.1 L of
Schiff reagent allows staining of 100 samples at a time. Once the stain is ready, samples
are submerged overnight (12 hours) in 1L fixative solution (85 methanol: 10 formalin: 5
acetic acid) and then rinsed in tepid running water for at least 10 minutes. This is
followed by a two hour hydrolysis treatment in 1L of 5.0 N HCL which depurinates
nucleotides and exposes aldehyde groups. During this time, Schiff reagent is decolourized by adding 8.25 g of activated charcoal to remove organic particles with a filter paper (#
40 Whatman) and a Buchner funnel. After hydrolysis is completed, samples are rinsed in
10 weak acid (1.0 N HCL) and treated for two hours with Schiff reagent. Schiff reagent
attaches to aldehydes and changes from colourless to pink.
After staining, excess stain is removed by three consecutive five minute rinses
with bisulfite solution (900 ml dH20, 50 ml 10% Na2S205, 50 ml 1.0 N HCl). All
samples are then rinsed for at least 10 min with running water followed by three rinses
with deionised water and dried gently with paper towel. All samples are stored in the dark
at room temperature until image analysis.
Image analysis is performed using a compound microscope (in this study, a Leica
LS DM) attached to a digital camera (in this study, an Optronics DEI-750 CE camera)
connected to a desktop computer. Before placing the samples on the compound
microscope, refractive oil is placed directly on the sample (no = 1.540) to eliminate the
cellular and nuclear membranes from sight. Then the sample is covered with a cover slip
and one drop of immersion lens oil is placed on top (no = 1.515) for use with a lOOx oil
immersion lens. Images are captured live with Bioquant Life Science Software (Version
8.12.20 MR for Windows XP; 2007 BIOQUANT Image Analysis Corporation, Nashville
TN) (Figure 1.4). Bioquant allows the measurement of optical densities (OD) on a linear
scale (0-250) under the green light channel of every pixel selected within a desired area
(stained cell nucleus) when compared to the background (clear area). In one step,
Bioquant provides an integrated optical density (IOD) for the whole nucleus that is used to calculate genome size estimates according to the following conversion:
GS (unknown) = GS (standard) x [mean IOD (unknown) / mean IOD (standard)]
Calculations for the standard species used in the present study are provided in Table 1.3.
11 Challenges in FIA estimates in insects
When estimating genome sizes using FIA, the major source of error comes from
measuring nuclei that differ in nuclear morphology (e.g., from different cell types) and
DNA compaction level (Hardie et al. 2002). In insects, this is often the result of nuclei
being damaged during dissections. Therefore, it is preferable to measure undamaged
nuclei with undisrupted borders from the same tissue type. Another problem observed in
insects is that many species store spermatozoa in tight bundles. Separation of
spermatozoa may be achieved by gently rubbing the dissecting pins against each other
during dissections, but some nuclei may be damaged in the process. As part of the preliminary work in the present study, I investigated whether the age of males in
Drosophila melanogaster Oregon R strain affects the nuclear DNA content in
spermatozoa. Genome size estimates did not differ significantly between adult males that were 1-2 days, 2-6 days, 1 week, 2 weeks, or 1 month old (ANOVA, F5; 12 = 1.12, p =
0.3999).
Flow cytometry
Flow cytometry (FCM) is a method for genome size estimation that applies
similar principles used in FIA. The amount of fluorochrome bound to nuclei is measured by estimating differences in light emissions (fluorescence) when passing through a laser beam or mercury arc lamp. This has been the second most common method for genome
size estimation in insects after Feulgen densitometry, though it is becoming increasingly used.
12 FCM in insects
The steps required to estimate genome size in insects using FCM are similar to
FIA. The key components in sample preparation are to co-prepare and co-stain the standard and the unknown samples. The preferred tissue used in this method is whole heads (usually 2C, with some 4C; 1C in male Hymenoptera) from adult insects and the most common standard used is Drosophila melanogaster heads (GS = 0.18pg) from female individuals. Samples are co-prepared and co-stained by first placing one head from the standard and the unknown in a 2ml Kontes Dounce grinder kept on ice containing 0.5 mL of Galbraith buffer (per 1L dH20: 8.8g of Na2C6H507,4.2g of 3-[N- morpholino]-propane sulfonate, 1.99g MgCL;, 1.0 mL Triton X-100, 100 uL lOmg/mL RNase A, adjusted to pH 7.2). Using a Type A pestle, the heads are ground together by gently moving the pestle up and down at least 15 times, then filtered through
30 uL nylon mesh (Spectramesh) into a 1.5 mL centrifuge tube (DeSalle et al. 2005).
The choice of dye used will depend mainly on the wavelength of the excitation source (laser beam) available in the flow cytometer. However, the most important factor is that the stain selected is a non-base-specific dye that will bind to all DNA found in nuclei. One of the most common dyes used for this purpose is propidium iodide (PI) which fluoresces when exposed to a light wavelength of 490 ran (Dole el et al. 2007).
To determine insect genome sizes, 25uL of PI stock solution (lmg/lmL) is added to each sample and incubated in ice for at least 20 minutes. Measurements are then taken and visualized in real time using a flow cytometer. Each sample takes 2-5 minutes in order to obtain > 1300 counts (Dole el et al. 2007) per peak required for reliable estimates. The coefficient of variation in insects is usually below 5%.
13 Once the data are collected, they must be analyzed using a procedure known as gating. Gating is the selection of a cluster of nuclei with similar DNA content and structure and elimination of debris (i.e. damaged nuclei) (Fig. 1.5). Genome size estimates are then calculated from the means of the 2C peaks of the standard and the unknown from the gated histogram (Fig. 1.5E). The conversion is similar to that used in
FIA:
GS (unknown) = GS (standard) x [mean 2C peak (unknown) / mean 2C peak (standard)].
Challenges of FCM estimates in insects
As with FIA, care must be taken to ensure accuracy of FCM measurements.
Preliminary methods development in the present study showed that the size of the insect, especially the head, can alter the resolution of the reading. The heads of large insects (e.g. crickets, grasshoppers, and dragonflies) contain large amounts of tissue other than brain, and substances such as mouth gland enzymes that may interfere with measurements (Fig.
1.6). Furthermore, insects such as the house cricket, Acheta domesticus, may contain nucleases that are liberated when the cells are ground and treated with detergent as in
FCM (Kaunelas 1970; Mangan and Kaunelas 1970). Such substances may be responsible for the absence of the D. melanogaster 2C peak when both are co-prepared (Figure 1.6A).
In addition, grinding of a large and small head together may make tissue extraction from the small head difficult since it may fit into grooves of the bigger head. To prevent this, it is best practice to dissect a small section of the brain of the large unknown and grind it together with the whole head of the small standard (Figure 1.6B). Small heads (e.g.
14 parasitoid wasps) may also be challenging since they may not provide enough counts (>
1300) for their peaks, but enough counts can usually be obtained if samples are run for
over five minutes.
Summary
The tremendous variation observed in the size of the genomes among animals
(more than 7,000 fold) has become a major puzzle in modern science. Genome size has
been shown not to vary randomly and is instead related positively to changes in cell size
and negatively to cell division rate. Despite the limited data available, it has been shown
that genome size is possibly linked via the cell to changes at the organism level including
body size, developmental rate, and flight activity.
The genome size data available is strongly biased toward vertebrates even though
most animals on the planet are insects. This gap in our knowledge imposes a major
obstacle to better understanding of the causes, mechanisms, and effects driving genome
size evolution. The lack of insect data may not be interpreted as a lack of interest in this
group, but this may be due to the difficulties measuring insect genome sizes until recently. Today, there are two accurate, and reliable techniques (FIA and FC) which
allow the rapid measurement of insect genomes. These methods were used in this study to
characterize the genomes of the insect orders Odonata (Chapter 2) and Hymenoptera
(Chapter 3).
15 Table 1.1. Currently known insect genome size diversity. This table shows the number of species analyzed (N), mean and standard error (M ± SE,
in pg), range (GS Range, in pg) and maximum and minimum ratio (Max / Min) for each order studied to date. Developmental mode (DM) for each
order is stated: ametabolous (AM), hemimetabolous (HE), and holometabolous (HO). This is based values reported until May, 2008 on the Animal
Genome Size Database (Gregory 2008), and does not include data reported in the present study.
Taxonomic classification N M±SE GS Range Max / Min DM
Insecta 471 1.48 ±0.12 0.11-16.93 154.18 -
Blattaria (Cockroaches) 4 2.78 ±0.29 2.00 - 3.36 1.58 HE
Coleoptera (Beetles) 177 0.66 ±0.04 0.16-5.02 31.38 HO
16 Diptera (True flies) 122 0.56 ±0.04 0.12-1.90 15.83 HO
Hemiptera (True bugs) 36 0.89±0.16 0.18-6.15 34.16 HE
Hymenoptera (Wasps, bees, and ants) 10 0.31 ±0.05 0.16-0.70 4.35 HO
Lepidoptera (Moths and butterflies) 53 0.65 ± 0.05 0.29-1.94 6.69 HO
Odonata (Dragonflies and damselflies) 14 1.12±0.18 0.37-2.16 5.84 HE
Orthoptera (Grasshoppers, crickets, and relatives) 43 8.98 ±0.55 1.55 -16.93 10.92 HE
Phasmatodea (Walking sticks) 9 3.02 ±0.63 2.00 - 8.06 5.03 HE
Strepsiptera (Twisted wing parasites) 2 0.12±0.01 0.11-0.13 0.85 HO
Thysanura (Silverfish and firebrats) 1 3.09 _ _ AM Table 1.2. General procedure in the Feulgen reaction for genome size estimation using FIA (adapted from Hardie et al. 2002).
Step Description
Prepare Schiff reagent: Make 1.1 L (1000mLdH2O, 11 g of basic fuchsin, 10 g of sodium metabisulfite, 100 mL 1.0NHC1) 36hrs
before hydrolysis to stain 100 samples. The stain should be dark purple.
Fixation: Fix samples by submerging them in 1L fixative solution (85 methanol: 10 formalin: 5 acetic acid) overnight.
Water rinse: Remove any excess fixative solution by rinsing samples with tap water for at least 10 minutes.
Hydrolysis: Treat samples with 1L of 5.0 N HC1 to depurinate DNA for 2 hours.
Decolourise Schiff reagent: During the hydrolysis step, add 8.25 g of activated charcoal to the Shiff reagent and filter through # 40 17 Whatman filter paper and a Buchner funnel. After this treatment, the stain should be clear or with a slight pink colour. If this is not the
case, repeat the step with half the amount of charcoal.
Weak acid rinse: Once the hydrolysis step is completed, quickly immerse samples in 1.0 N HC1.
Staining: Treat samples with decolourised Schiff reagent for 2 hours.
Bisulfite rinse: Remove any excess stain by rinsing samples with 3 five minute changes of bisulfite solution (900 ml dH20, 50 mL
10% Na2S205, 50 mL 1.0 N HCl).
Water rinse: Remove any excess bisulfite solution by rinsing samples with deionised water for at least 10 minutes.
10 Drying: Dry samples gently with paper towel and store in the dark at room temperature. Table 1.3. FIA mean Integrated Optical Densities (IODs), coefficients of variation (CV, in %) and haploid genome size estimates (GS, in pg) obtained in 3 different Feulgen reactions (FR) for the internal standards and insect standards. Erythrocytes from Gallus domesticus (chicken, GS =
1.25 pg; Rasch et al. 1971; Hillier et al. 2004) and Oncorhynchus mykiss (rainbow trout, GS = 2.60 pg; Rasch 1985; Dolezel et al. 2003), and spermatozoa from the insect standards Drosophila melanogaster Oregon R strain (vinegar fly, GS = 0.18 pg; Rasch 1971) and Tenebrio molitor
(yellow mealworm beetle, GS = 0.52 pg; Juan and Petitpierre 1989). FIA methodology follows that of Hardie et al. (2002).
FR1 FR2 FR3 Mean GS ± SE "
Species IOD CV GS IOD CV GS IOD CV GS
G.domesticus 481.74 2.54 1.25 455.21 2.54 1.25 455.29 2.74 1.25 1.25
O. mykiss 983.20 2.61 2.55 912.11 2.50 2.50 934.50 2.57 2.57 2.54 ± 0.02
D. melanogaster 39.84 6.78 0.18 44.00 6.63 0.18 45.19 5.04 0.18 0.18
T. molitor 110.61 5.16 0.50 128.80 4.31 0.53 132.64 5.83 0.53 0.52 ±0.01 Vertebrate* _ 3% 1... A
B
Figure 1.1. A) Diversity of named species of animal groups (Grimaldi and Engel 2005). This is not reflected in genome size studies. B) Distribution of genome size estimates in animals (n =
4475) showing percentages for vertebrates (n = 2995), insects (n = 471), and other invertebrates
(n = 1009). As of May 3rd, 2008 based on the Animal Genome Size Database (Gregory 2008).
19 30 27% 25% 25 22% co •4—i CO •111 18% llliBBiB§K o SHlllllllll Zi Q_ 15
£ 10 ZJ 6% jfflHiiflHB mmSMBffl& 2% mmsm Blilll 0 60s 60s 70s 80s 90s 2000 to 2008
Figure 1.2. Number and percentage of publications (total = 96) reporting insect genome size
estimates since the first published estimate in 1952. As of May 3rd, 2008 based on the Animal
Genome Size Database (Gregory 2008).
20 11=471 : 200 -
150 - m -
*t 50 • 1 1 1 ,M 1 • 0 4 ""f 'i I "'!'" "«• t n ? j" j"- 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 GS (pg)
Figure 1.3. Genome size (GS) variation in insects (n = 471 species). As of May 3rd, 2008 based on the Animal Genome Size Database (Gregory 2008).
21 j i, oil „*£ oti*HO« *. 4 A <3ta,®m.omm®t& BO «fi
jj t ht Current A £ri'7 7 Figure 1.4. Genome size estimation using FIA and the Bioquant Life Science Software (Version 8.12.20 MR). Pixel value threshold for nuclei is selected under green light and compared to that of the background (clear area) to obtain the Integrated Optical Density (IOD) for each nucleus. 22 Figure 1.5. Genome size estimation using flow cytometry. Gating of a cluster of nuclei with similar DNA content. For measurement, the integrated fluorescence parameter FL2-Area was set for linear correlation with the amount of DNA content. First, nuclei were selected from a scatter plot of FL2 height (FL2-H) vs FL3-H (A). Second, a scatter plot of side scatter height (SSC-H) vs FL2-Area (FL2-A) was used to gate the standard and the unknown 2C clusters (B). Third, gated clusters were displayed as peaks on a histogram of FL2-A vs nuclei counts (C). An ungated histogram (containing all data points) FL2-A vs nuclei counts was also used to ensure that data were gated properly (D). Data summaries for each peak from the gated histogram were displayed on a table showing the number of nuclei, gating percentage, mean, coefficient of variation, and peak maximum (E). Genome size estimates were calculated from the means of the 2C peaks of the standard and the unknown using the conversion formula shown (F). Measurements were performed on a BD Biosciences FACSCalibur flow cytometer (BD Biosciences, San Jose, CA). Data were obtained and analyzed using the software Cell Quest Pro (Becton Dickinson and Co., 1996). Sample preparation is explained in detail in the text. 23 Gated Marker Events % Gated Mean CV PeakCh All 13939 100.00 528.8? • SS.69 SS Standard 2C 393? 28,24 59.82 S.36 58 Unknown 2C 10002 71,78 713.21 1.96 714 Standard 4C 0 0.00 Unknown 4C 0 0.00 Genome Size Calculation: C-value{U) == C-value (S) X Mean (U) Mean (S) C-value{U) == 0,18 pg X 713.21 59.82 C-value {Unknown} = 2.15 pg 24 A 1*00 »i Sj Standard 2C *§. S Unknown W «s*iw*nj8c B *o. 1 I f%«**~^S Mandative n,i^„,,^iri,vm,^,!r^r • f^*f"i™p^^pi< » >•;•»» friummmn* ?00 Figure 1.6. Comparison of FCM measurements between the whole head (A) and the isolated brain (B) of Acheta domesticus (house cricket). In both cases, the samples were co-prepared with one head of Drosophila melanogaster Oregon R strain female as the standard. The standard peak is not present when the whole head of the house cricket is used (A). The peaks shown are ungated. 25 Chapter 2: Genome size diversity in relation to flight strategies in the insect order Odonata (dragonflies and damselflies) 26 Abstract Odonates (dragonflies and damselflies) are one of the most charismatic and best characterized insect groups on Earth. The manageable species diversity, large body size variations, and exceptional flying abilities make them an ideal insect group for examining the evolution of genome size and its possible relation to these and other parameters. This study presents the first large-scale attempt to document and analyse odonate genome sizes by providing estimates for 100 North American species. Genome sizes did not correlate with body size, voltinism, or nymphal habitat, but showed a weak positive correlation to chromosome number. Interestingly, odonate genomes were found to be larger in fliers than in perchers. This suggests, for the first time in insects, a possible link between genome size and flight capability. 27 Introduction The Odonata' are among the most recognizable insects alive today. This has undoubtedly facilitated their study, allowing them to become one of the best documented insect orders in North America and the world (Westfall and May 1996, Needham et al. 2000). The order Odonata is divided into three suborders: Anisoptera (dragonflies), Zygoptera (damselflies), and Anisozygoptera. Worldwide, ~6,000 species of Odonata are recognized, classified into 33 families, including 11 dragonfly families, 21 damselfly families, and one Anisozygoptera family (Schorr et al. 2008). In North America, there are seven dragonfly families and eight damselfly families comprising a total of 96 genera and 511 recorded species (Westfall and May 1996, Needham et al. 2000) (Table 2.1). The strong state of knowledge about the diversity and biology of odonates makes them excellent subjects for the study of genome size evolution in insects. Hence, the two main objectives in this study were to significantly increase the number of species estimates for this group, and more importantly, to use the new data to examine the relationships, if any, between genome size and other biological parameters including developmental rate, body size, chromosome number, and flight ability. The hypothesized links between each of these parameters and genome size are presented on Table 2.2 and are based on the strong relationships established in previous studies between genome size and cell size (positive), and genome size and cell division rate (negative) (Bennett 1971, 1972; Cavalier-Smith 1978,1982; Olmo 1983; Gregory 2001, 2005). The name "Odonata" refers to their large and numerous teeth that allow them to be voracious predators of other insects (Gr. odonto, tooth). 28 Genome size diversity There have been few attempts to characterize genome size diversity in odonates. The Animal Genome Size Database (Gregory 2008) lists only 14 genome size estimates for this order, ranging about 5-fold from 0.37 pg (Gomphus flavipes and Ophiogomphus cecilia) to 2.16 pg (Rhinoaeschna confusa), with a mean value of 1.12 ± 0.18 pg (Table 2.3). Odonates exhibit some of the smallest genomes among hemimetabolous insects, with genome sizes mainly below the hypothetical 2 pg upper threshold (Gregory 2002b, 2005; Chapter 1). This suggests that odonate genomes may be under selection for attaining and/or maintaining small genomes. As found in some other arthropods, genome size constraints may be imposed by body size restrictions (e.g., copepod crustaceans; Gregory et al. 2000) or developmental time and complexity constraints similar to those found in holometabolous insects (Gregory 2002, 2005) (Table 2.2). Another possibility, one that is explored in detail in this chapter, is that energetic constraints imposed by flight, similar to those observed in birds (Hughes and Hughes 1995, Gregory 2002a), may play a dominant role in shaping genome size diversity among dragonflies and damselflies (Table 2.2). Body size diversity Adult body size in Odonata is primarily determined during their nymphal stage (Rowe and Ludwig 1991). In general, dragonflies are larger than damselflies, but there are significant differences in size at the family and genus levels. Some of the largest odonates are found in the family Aeshnidae, which are about 70mm long, while some 29 damselfly species such as Nehalennia Irene (Cordulegastridae) may be only 25mm in length (Westfall and May 1996, Needham et al. 2000). In animals, body size may increase as a result of higher cell numbers or by increasing cell size (Gregory 2000). If genome size affects cell size (as in the nucleotypic theory; Bennett 1972, Gregory 2001, 2005a) and cell numbers do not change greatly, then there can be a link between DNA content and body size. On the other hand, extensive changes in cell number may mask any effect of differences in cell size, as appears to be the case in many vertebrates (Gregory 2002c, 2005; Table 2.2). The hypothetical relationship between genome size and body size has been tested before in insects such as moths (Gregory and Hebert 2003) and ladybird beetles (family Coccinelidae) (Gregory et al. 2003), and these studies did not find any significant correlations. However, these studies were of limited scope, and it remains a possibility that these two parameters are related in other insects such as odonates which range greatly in body size. Developmental rate (voltinism) Odonates are hemimetabolous, exhibiting incomplete metamorphosis. They are aquatic during their nymphal stages and terrestrial during their adult stage. Nymphal stages may last up to five years, whereas the adult stage usually lasts between one and three months (Corbet 1962, 1999, Corbet et al. 2006). The number of life cycles per year (voltinism) has been a major area of study in this group. Although voltinism depends to a significant extent on environmental factors such as rainfall and temperature, and despite some notable intraspecific variation, odonates have been classified as multivoltine (> 3 30 generation/year), bivoltine (2 generation/year), imivoltine (1 generation/year), semivoltine (1 generation/ 2 years), and partivoltine (1 generation/ > 2 years) (Table 2.4) (Corbet et al. 2006). No detailed data are available on temperature-controlled developmental rates in odonates as they are for some other insects (e.g., Drosophila species; Gregory and Johnston 2008), but voltinism can be taken as a rough approximation of how quickly different species complete their ontogeny. Corbet et al. (2006) found evidence for a correlation between voltinism and nymphal habitat between species. Specifically, three types of odonate habitats were recognized: 1) temporary waters, 2) perennial lentic waters (moving waters including streams and rivers), and 3) perennial lotic waters (stationary waters including ponds, wetlands, and lakes). Species that develop in temporary waters usually have a generation time of one year or less. Perennial lentic waters have similar proportions of species with all voltinism modes, while species living in perennial lotic waters are mainly univoltine, semivoltine, or partivoltine (Corbet et al. 2006). In other words, species with faster development rates are most often found in temporary waters whereas species with slower life cycles are most often found in perennial waters. As a result of associations between larger genome size and slower cell division (Gregory 2001, 2005), developmental rate may correlate with genome size in odonates as it does in some other insects such as beetles (Gregory et al. 2003) and flies (Gregory and Johnston 2008) (Table 2.2). 31 Flight strategies Corbet (1962) divided the Odonata into "fliers," "perchers," and "gliders" (Table 2.5). He defined fliers as "those which, when active, remain constantly on the wing," perchers as "those which spend most of the active period on a perch from which they make short flights," and "gliders" as those which have "a hyperdevelopment of the anal field of the hindwing;" which "enables them to glide during sunshine and thus to remain airborne at the expense of minimum activity of the wing muscles." An important distinction between fliers and perchers is that the former are considered to be endotherms while the latter are ectotherms (Corbet 1962, 1999; Heinrich 1974; May 1976, Heinrich and Casey 1978; May 1981,1991; Sformo et al. 2006). All damselflies are perchers, and the only perchers among dragonflies are the members of the families Gomphiidae and Libellulidae. All fliers belong to the dragonfly families Aeshnidae, Cordulegastridae, Corduliidae, and Macromiidae. Almost all gliders are migrant libellulids that also show a percher body type but are most often observed flapping their wings for short periods and taking advantage of wind currents to glide and travel (Corbet 1962). From the few gliders described to date, three are included in this study: Tramea Carolina, T. lacerata, and Myathiria marcella (Corbet 1962, 1999; May 1981). Associations between metabolic rate and genome size (via cell size effects on gas exchange) have been reported in mammals and birds, with the effect most pronounced in taxa that exhibit powered flight (i.e., birds and bats) (Szarski 1970; Burton et al. 1989; Hughes and Hughes 1995; Hughes 1999; Gregory 2002a, 2005). Even though flight first evolved in the Insecta and the vast majority of flying animals are members of this group, this relationship has not yet been examined within any insects. The gas exchange system 32 of insects differs fundamentally from that of vertebrates, but effects on other cell types (e.g., muscles) may still result in energetically important relationships between genome size and cell size. Based on a straightforward assessment of Corbet's classification of odonate flight strategies (Corbet 1962, 1996), it might be expected that constantly active fliers have the highest metabolic demands and therefore the smallest genomes, followed by more frequently stationary perchers and gliders with larger genomes. However, a closer examination of the biology of these insects reveals a more complex situation. First, fliers control body temperature and prevent heat loss from their thoracic flight muscles with surrounding air sacs located under the cuticle (May 1981). Second, they are generally large and have small flight muscles (25% body weight) which enhances insulation by allowing more space for air sacs (May 1981, 1991). Third, their flight muscles can generate enough heat to keep their bodies warm during constant flight even at low ambient temperatures (Sformo et al. 2006). Fourth, fliers rarely fly at their highest capacity (which allows rapid acceleration and manoeuvrability in flight) (Riippell 1989). Finally, fliers often take advantage of air currents for long periods of time (May 1981, 1991; Grabow and Rueppell 1995; Wakeling and Ellington 1997; Corbet 1962, 1999; Thomas et al. 2004). Perchers, by contrast, have large thoraces with extensive musculature (45% body mass), lack subcuticular air sacs, and most often fly at their maximum capability (May 1981, Riippell 1989; May 1991; Grabow and Riippell 1995; Wakeling and Ellington 1997; Corbet 1999; Thomas et al. 2004). In addition, they are usually small, which increases their surface area to volume ratio, thereby making the loss of heat more severe 33 (May 1981); moreover, they generally lack the ability to generate enough heat with their thoracic muscles to maintain a constant body temperature during flight (Sformo et al. 2006). To counteract the continuous loss of heat to the environment, perchers lay on hot surfaces, usually under direct sunlight, and modify their body posture to control their internal temperature (Corbet 1962, 1999; May 1976, 1981). At rest, they must maintain their body temperature high enough to take off at any given time to avoid predators, defend their territories, and pursue prey or mates (Corbet 1962). Therefore, perchers may have to maintain a higher metabolism than fliers even at rest. Perchers also seem to have a considerable disadvantage in terms of energy efficiency with regards to their overall wing design. Fliers have evolved long and narrow wings that may facilitate flight in open areas at higher speeds and over longer periods of time while reducing energy consumption (May 1976, 1981, Ruppell 1989, Grabow and Ruppell 1995). In contrast, perchers have short and wide wings that may be advantageous to manoeuvre at lower speeds, and be more suitable for flying through branches and trees (May 1976, 1981, Ruppell 1989, Grabow and Ruppell 1995), even though this maybe more energetically costly. The body design of gliders is very similar to that of perchers but their behaviour is similar to that of fliers, presumably resulting in very active but low efficiency flying with the highest metabolic demands. This suggests that gliders may have developed a better ability to maintain food reserves and/or be more efficient at using their energy with maximum output to be able to sustain continuous flight. In short, fliers are the most active, but this is analogous to efficient, long-distance walking; perchers and especially gliders are most similar to sprinters. 34 Cytogenetics There has been significant discussion regarding chromosome-level evolution in the Odonata (Oguma 1930; Dasgupta 1957; Cumming 1964; Kiauta 1967, 1968, 1969a; Agopian and Mola 1988; Prasad and Thomas 1992; Mola and Papeschi 1994). Karyotypes in this order are composed of holokinetic chromosomes that can range in haploid number from n = 3 to n = 15 (Cumming 1964; Cruden 1968; Kiauta 1972). However, more than 90% of the species examined to date possess 13-15 chromosomes. Odonate karyotypes often include microchromosomes (m-chromosomes) and are usually reported for haploid sets as Nm, in which N equals the total number of chromosomes and m equals the presence of one microchromosome in the total chromosomal set (Cumming 1964; Cruden 1968; Kiauta 1972). The sex determination system in the order is XX$: XOc?, but a few exceptional cases show an XX?: XY$ system (Kiauta 1979; Perepelov et al. 1998; Ray-Chaudhuri and Dasgupta 1949; Seshachar and Bagga 1962). Odonate m-chromosomes do not segregate like other autosomes during meiosis, and their origin and role is unclear. Early discussions suggested that microchromosomes are the remnants of an autosome in the process of elimination by progressive loss of chromatin (Oguma 1930). Oguma (1930) suggested that the "ancestral" Odonata karyotype consisted of 13 autosomes and a sex chromosome (n = 13a + X). From this starting set, an autosome evolved into an m-chromosome (n = 12a + m + X) that would eventually be lost (n = 12a + X). This process repeated a number of times resulting in the type number observed in the dragonfly families Petaluriidae (n = 9m) and Gomphiidae (n = 12orl2m). 35 However, Cumming (1964) partially rejected Oguma's (1930) chromatin loss hypothesis and stated that progressive loss may be an ongoing process, albeit one too slow to account for the gross chromatin rearrangements observed in Odonata. Instead, he suggested that such reductions could only be achieved by chromosomal fusions. Shortly after this, Kiauta (1967, 1968, 1969a, 1969b) made some modifications to Cumming's (1964) hypothesis while completely rejecting Oguma's (1930) hypothesis. Kiauta (1967) suggested that the type number for all Odonata was equal to 9 based on the hypothesis that the family Petaluriidae was the most "basal" extant dragonfly lineage and that any variation in this number was caused by random fusions and fissions. However, the assumption that the chromosomes of modern petaluriids are the same as the ancestral karyotype is potentially problematic; their current karyotype may be of secondary origin as Kiauta (1969b) suggested for the family Gomphiidae. It is interesting that odonate chromosomes appear to lack heterochromatin almost entirely (Prasad and Thomas 1992). C-band patterns in odonate chromosomes are not only weak, small, and difficult to demonstrate, but they are restricted exclusively to telomeres of autosomes (Prasad and Thomas 1992, Perepelov et al. 1998, Perepelov et al. 2001). This may be suggestive of DNA loss, but hypotheses regarding loss versus fusion can be most effectively tested with data on DNA contents. To this end, the present study includes an evaluation of whether chromosome number and genome size are correlated in odonates (Table 2.2). 36 Materials and Methods Collection, identification, and voucherins of specimens Odonates were collected in two main geographic regions: Ontario, Canada and Florida, USA. Collection in Ontario took place between June and September 2006 and 2007, primarily around Guelph and Algonquin Park. Florida samples were collected during May 2007 in Tallahassee (northern Florida) and at the Archbold Research Station (Lake Placid, central Florida). Most dragonflies and damselflies were mature adult males hand-collected using nets. In total, 427 specimens were included in this study. Taxonomic identifications were provided by Colin Jones (Ontario Ministry of Natural Resources), Nancy Deyrup (Archbold Research Station), and Michael May (The State University of New Jersey). All specimens were stored in cellophane envelopes at room temperature in the dark. The specimens were not treated with any chemicals (e.g. acetone) for preservation since these samples were donated to the Biodiversity Institute of Ontario (Guelph, ON) for DNA barcoding (Hebert et al., 2003). Specimens were photographed and images and collection information were deposited in the Barcode of Life Data Systems (BOLD) (http://www.boldsystems.org). Each specimen is in the process of being barcoded by staff at the Canadian Centre for DNA Barcoding and this information eventually will be available online in order to facilitate future identifications of the same species and future verification of the identities of the specimens used in this study. Genome size estimation Spermatozoa represented the primary cell type analyzed in this study. One of the main reasons for this is that males are considerably easier to obtain than females in the 37 field. In addition, mature males may contain thousands of spermatozoa, often found tightly packed in sacs known as spermatodesmata. Males may have hundreds of spermatodesmata in their vas deferens located in the 9th abdominal segment (Abro 1999, 2000). Additionally, their nuclei stain uniformly with Schiff reagent in Feulgen reactions (Seshachar and Bagga 1963; Chapter 1). Odonate genome size estimates were obtained using a Feulgen image analysis (FIA) protocol modified from Hardie et al. (2002). Testes were dissected from adult males in Ringer's saline and placed in solution on a glass slide. Odonate spermatozoa were often found in tight bundles, and were separated from each other by gently rubbing a pair of pins against each other. Samples were then allowed to air dry at room temperature and stored in the dark until staining in the Feulgen reaction (Chapter 1). In a small number of cases (n = 7), sperm could not be obtained and haemocytes were used instead. Haemocytes were also measured in five individuals from each of four randomly selected species from which spermatozoa could be obtained to compare results with genome estimates obtained from spermatozoa (Table 2.6). Genome size estimates from haemocytes and spermatozoa did not differ significantly for these species (paired t-test, p = 0.32). Body size and wing area measurements Odonate specimens were weighed (to the nearest 0.001 g) using a Denver M-120 analytical balance. Measurements were taken at least six months after collection, but dry weights per se were not measured since this requires an intensive heat treatment of 60 °C for 24 hrs (Johnston and Cunjak 1999), a procedure which may inhibit DNA extraction for DNA barcoding analysis and other future genetic work. 38 Body size (length) and wing size (length and area) were measured for all specimens using photographic image analysis (PIA). Digital images of specimens were taken using a Canon 30D digital camera with a 100mm macro lens set up on a Polaroid MP-4 Land Camera. A 10 mm2 section of millimetre-squared graph paper was placed in between two glass cover slips held together with two small pieces of tape. This was used as the scale for length and area in every picture. Details of the camera settings, setup and procedure are provided in Appendix 1. Each image was uploaded into NIS-Elements BR software 2.30 (Laboratory Imaging, Nikon, 1991-2007) and the length, in pixels, was measured first on the scale bar (10 mm). Then, length from front to back was measured for the head, thorax, abdomen, one forewing and one hindwing (Fig. 2.1). Pixels were converted to millimetres in MS Excel (Microsoft Corporation, Redmond, Washington, USA). Total body length was obtained by summing the measurements for head, thorax, and abdomen. Often, abdomens were broken in two or more parts. In these cases, each part was measured separately and summed. Wing lengths and areas were measured from the base to the most distant edge of the tip (Fig. 2.1). Wings that were damaged or folded were not measured. In most cases this did not impede measurements, since usually at least one of each of the forewings and hindwings was undamaged and in good condition. The wing loading index (WLI) (in g / mm ) was calculated for each species by dividing minimum mass by the total surface area of all four wings. Voltinism and habitat Voltinism and habitat for 36 of the species in this study were obtained from the most complete and recent review on the subject by Corbet et al. (2006). Voltinism codes 39 were assigned based on the fastest life cycle recorded for each species: 1 = multivoltine, 2 = bivoltine, 3 = univoltine, 4 = semivoltine, and 5 = partivoltine. Each species was also classified according to habitat: 1 = temporal waters, 2 = perennial lentic waters, and 3 = lotic perennial waters (Corbet et al. 2006). Karyotype data Karyotypes were compiled from the literature for 54 species (34 dragonflies and 17 damselflies) obtained mainly from two studies that characterized North American odonates (Cumming 1964; Cruden 1968) (Table 2.7). Statistical Analysis In this study, genome size estimates were presented as the mean ± standard error for each species. The relationship between genome size and body size was examined using a Pearson correlation test while the relationships between genome size and voltinism, habitat, and flight strategy were examined using an analysis of variance (ANOVA). The degrees of phylogenetic conservatism at the sub-order, family, genus, and species levels were assessed using a nested ANOVA. Lastly, the relationship between genome size and chromosome number was analysed using a Spearman's rank correlation. Results Genome size diversity Based on the 100 species characterized in this study (Table 2.7), odonate genome sizes range about 6-fold, from 0.41 pg in Myathiria marcella to 2.38 pg in Somatochlora elongata. The mean value for the order is 1.01 ± 0.04 pg. Dragonfly and damselfly mean values were 1.01 ± 0.05 pg and 1.02 ± 0.04 pg, respectively. Previous family and genus 40 means and ranges are concordant with estimates provided in this study except within the family Gomphiidae for which previous estimates were roughly half (0.37 - 0.40 pg) of what was reported here (0.54 - 0.94 pg). Estimates from earlier studies were not included in other analyses for the sake of consistency. A nested ANOVA estimated that 0.0% of the variation in the dataset occurs at the level of suborders within the order, 61.1% among families within the two suborders, 30.8% across genera within families, and 8.1% among species within genera. In this study, the family Gomphiidae spanned a genome size range from 0.57 pg (Dromogomphus spinosus) to 0.94 pg (Ophiogomphus rupinsulensis). In Libellulidae, the large range observed (GS range = 0.41 - 1.27 pg) is caused mainly by the tropical dragonfly Miathyria marcella (GS = 0.41 pg) and members of the genus Leucorrhinia (GS > 0.94 pg). Similarly, the large variation observed in Corduliidae (GS range = 0.98 - 2.36 pg) is due to the large genomes reported here for the genera Cordulia (GS = 1.54 pg) and Somatochlora (GS > 1.80 pg). Lastly, within Aeshnidae all genomes larger than 1.50 pg are found in the genus Aeshna (Table 2.7). The majority of the damselfly estimates reported here (74%) belong to the family Coenagrionidae and range from 0.88 to 1.80 pg (Table 2.7). Most genomes in this family were below 1.35 pg with the exception of Nehalennia integricollis (GS = 1.53 pg) and N. irene (GS = 1.80 pg) (Table 2.7). Body size and wins area relationships Species measurements reported here for total length (TL), abdomen length (AL), and hindwing length (HWL) are within the ranges reported previously (Westfall and May 1996, Needham et al. 2000). In addition, this study provided mean values for 41 minimum total weight (TW), head length (HL), thorax length (TL), forewing length (FWL), forewing area (FWA), and hindwing area (HWA) (Table 2.8). In general, values for the various morphometric parameters were strongly correlated (Appendix 2), such that any of them could provide a useful indicator of body size among species. These correlations remained positive and highly significant (usually r > 0.900 andp < 0.001) at the order and suborder levels (Appendix 2). Morphological parameters did, in some cases, vary somewhat independently of one another. Notably, in the damselfly family Coenagrionidae, wing length and area measurements were positively correlated with each other and with abdomen length (r > 0.99 andp < 0.05), but not as strongly with other indicators of body size (Appendix 2), perhaps indicating specializations involving substantial increases of wing size relative to body size. Genome size vs body size Genome size showed a positive and significant correlation with TL (r = 0.24, p <0.02, n = 97) and AL (r = 0.32,p < 0.002, n =.97) across all odonate species but not with any other body size parameter (r < 0.12 and/) > 0.22, n = 97). The same pattern was observed at the genus level, however, no correlations were significant at the family level (r < 0.32 and/? > 0.40, n = 9) (Appendix 2). In the dragonfly suborder nearly all body size parameters showed positive significant correlations to genome size at the species and genus levels (r > 0.30, p < 0.03) with the only exception found in THL at the genus level, where this was marginal (r = 0.30, p < 0.09). However, no correlates were significant at the family level (r < 0.78, p > 0.06) (Appendix 2). Within damselflies all body size relationships with genome size were 42 found to be negative but they were significant only at the species level (r < D0.35 andp < 0.03) (Appendix 2). Despite the significant correlations found at the species and genus levels, close examination suggests that this relationship is of limited scope in these insects. Notably, whereas some large odonates do exhibit large genomes - for example, Aeshna canadensis (TL = 67.58 mm; GS = 2.20 pg), Cordulia shurtefli (TL = 43.68 mm; GS = 1.54 pg), and Somatochlora. williamsoni (TL = 49.08 mm; GS = 1.80 pg) - small body size clearly does not require small genome size as some of the smallest species displayed some of the largest genomes, as seen in Nehalennia integricollis (TL = 22.82 mm; GS = 1.53 pg) and N. Irene (TL = 22.96 mm; GS= 1.80 pg). Additionally, species with very similar genome sizes can differ greatly in body size, as observed in the gomphids Hagenius brevistylus (TL = 73.08 mm; GS = 0.93 pg) and Ophiogomphus rupinsulensis (TL = 49.30 mm; GS = 0.94 pg), and the coenagrionids Argia moesta (TL = 23.87 mm; GS = 0.89 pg) and Amphiagrion saucium (TL = 40.89 mm; GS = 0.90 pg). Odonate genomes vs nymphal habitat and voltinism Genome size did not differ significantly among nymphal habitats at the species level across the order (ANOVA, F2,33 = 1.12, p = 0.6237) or in the dragonfly (ANOVA, F2,io = 0.24,p = 0.7873) and damselfly (ANOVA, F2,13 = 0.43,p = 0.6600) suborders. The same was observed at the genus level across the order (ANOVA, F2,19 = 0.15, p = 0.8579). However, genome size and voltinism were related positively and significantly at the species level in the order (ANOVA, F3) 32 = 3.79,p = 0.0198) and among dragonflies (ANOVA, F3,19 = 3.79,p = 0.0275) but not among damselfies (ANOVA, F2,10 = 1.01,/? = 0.3978). 43 Odonata genomes vs karyotype In dragonflies, the lowest chromosome number is found in the family Gomphiidae (n = 12m or 12), followed by the family Libellulidae (n = 13m or 13), and the family Aeshnidae (n = 14m or 14). In damselflies a similar pattern is observed, in which the families Lestidae (GS = 0.64pg ± 0.02) and Calopterygidae (GS = 1.03pg ± 0.04) overall share the same chromosome numbers (n = 13m or 13), while members of the family Coenagrionidae have larger genomes (GS = l.llpg ± 0.04) and more chromosomes (n = 14m or 14). Spearman rank correlations between genome size and chromosome number of 54 species showed a significant positive correlation (rs = 0.41, p = 0.002). Similar results were obtained for the damselfly suborder (rs = 0.49, p < 0.05, n = 17) but in dragonflies the relationship was not significant (rs = 0.23, p = 0.15, n = 34). In two species of the dragonfly family Corduliidae, chromosome numbers were considerably smaller than the type number (n = 13m) for the family and their genome estimates were usually below the mean value (GS = 1.35 pg) for the family: Dorocordulia libera (GS = 0.98pg; n = 6 or 7) E. cynosura (GS = l.lOpg; n = 10 or 11) (Cruden, 1968). Odonata senomes vs flight strategies There was a significant difference in genome size estimates between fliers (all of which are dragonflies), percher dragonflies, percher damselflies, gliders and the percher damselflies N. irene and N. integricolis ("hoppers": remain in small dense areas, their activity may be limited to very short flights and hopping between branches mainly) (ANOVA, F4,95 = 28.4,;? < 0.0001). A Fisher LSD test revealed three groups in which mean estimates differed significantly: a) hoppers (GS = 1.66 ± 0. 14) and fliers (GS = 44 1.43 ± 0. 09), b) percher damselflies (GS = 0.98 ± 0.04), and c) percher dragonflies (GS = 0.78 ± 0.02) and gliders (GS = 0.58 ± 0.03) (Fig, 2.2). The mean wing loading index (WLI) differed significantly between the flight groups (ANOVA, F4,92 = 18.2, p < 0.0001). A Fisher LSD test showed that fliers (WLI = 0.114 mg/mm2, n = 23) posses the highest WLI and differed significantly from all other groups. Percher dragonflies (WLI = 0.090 mg/mm2, n = 35) possessed a WLI that was significantly larger than that of percher damselflies (WLI = 0.051 mg/mm , n = 34), gliders (WLI = 0.047 mg/mm2, n = 3), and hoppers (WLI = 0.028 mg/mm2, n = 2). There were no significant differences in the latter three groups. Lastly, there were no significant mass-corrected correlations (all/? > 0.1272) found between wing-loading index and genome size at the species, genus, or family levels in the order or in either suborder and between the flight groups. Discussion Genome size diversity The 100 genome size estimates generated in this study support previous suggestions that odonate genomes are small to medium in size (0.41-2.38 pg) and appear to be constrained relative to other hemimetabolous insect orders. This may be explained by the fact that odonates have a larval aquatic stage (unlike most other hemimetabolous insects) in which they undergo major morphological changes during their second last and last molts to attain the characteristics of their terrestrial forms. This process may not be as drastic as the pupal stage in holometabolous insects, but it may resemble the constraints imposed the size of the genome during this process. 45 In general, odonate genomes tend be consistent in size within genera and families and variation within species was usually below O.lpg with a standard error below ±0.05 (Table 2.7). For the most part, new and previous estimates agree, with the only exception found in genomes from the family Gomphiidae. Previous estimates for three Eurasian species of this family (Petrov and Aleshin 1983) are roughly half of those reported here for 14 other species (mean of 0.38pg vs. 0.76pg). This difference may suggest that there is considerable genome size variation between gomphiid species that are located in different geographical regions, though it simply could be the result of error. Body size Overall, genome size was not convincingly related to body size in odonates. While some relationships were observed, it is worth noting that some of the largest genomes reported here are found in some of the smallest odonates. This perhaps is not surprising given that body size in adults is determined during the nymphal stage and is affected by environmental factors such as temperature and diet (Rowe and Ludwing 1991). In addition, there is high inter- and intra-specific body size variation in adult odonates (Bried et al. 2005; Bried and Ervin 2007). It also suggests that cell number, rather than cell size, may be a primary determinant of body size in odonates, unlike in copepod crustaceans with cell number constancy resulting in a genome size-body size correlation (Gregory et al. 2000). Nymphal habitat and voltinism Similar to body size, developmental rate did not appear strongly linked to genome size in most taxonomic groups. Significant correlations were found across the order and among dragonflies, but these may be of limited applicability given that 25 of the 36 46 species examined are univoltine and their genome sizes ranged from 0.54-1.44 pg. The only five semivoltine and two partivoltine species included in this study showed genome sizes that ranged between 0.74 and 2.00 pg and 0.93 and 1.54 pg, respectively. In addition, nearly all bivoltine species considered in this study have been reported also as univoltine in some environments and their genomes ranged widely, varying from 0.56 to 1.20 pg. Odonate nymphal habitats and voltinism modes are highly variable between and within species. Many species can change their developmental rates depending on water availability, temperature, and latitude (Corbet et al. 2006). For instance, a tropical species may switch to a different voltinism mode if found in temperate regions (Corbet et al. 2006). In addition, a major portion of the 275 species reported by Corbet et al. (2006), with more than one voltinism record, were reported with different voltinism modes. The only exception to this in Odonata may be found in the genus Lestes, which has the smallest genome sizes among damselflies and is the only group considered to be composed only by obligate univoltines (Corbet et al. 2006). The flexibility in developmental rate and lack of a significant relationship with genome size suggest that developmental rate does not impose major restrictions to genome size in odonates as it appears in holometabolous insects (Gregory 2002a, 2005; Gregory and Johnston 2008). Holometabolous insect orders seem to have a roughly 2 pg upper limit (Gregory 2002a, 2005), but hemitabolous insects orders including Odonata, may have species with genomes that exceed this limit. 47 Cytogenetics Overall, chromosomal number and genome size correlate positively in odonates. This suggests that changes in the karyotypic arrangement of odonate genomes by loss or gain of chromosomes are linked to decreases and increases in DNA content. Members of the family Gomphiidae have fewer chromosomes (n = 12m) and smaller genomes on average (GS = 0.76 ± 0.03 pg) than any other dragonfly family, whereas members of the family Aeshnidae have both more chromosomes (n = 14m) and larger genomes (GS = 1.61 ± 0.10 pg). A similar pattern is observed among damselflies in which coenagriionids have the highest chromosome number (n = 14m) and the larger genomes in this suborder (1.11 ±0.04). Reductions in chromosome numbers do not necessarily mean reductions in genome size. Cumming (1964) reported that the dragonfly Macrothemis hemiclora contained three chromosomes and not 13 as other species of the same genus, but their chromosomes seemed to contain the same amount of DNA. A similar conclusion is attained from the genome estimates generated in this study and the karyotypes previously reported by Cruden (1968) for three Epitheca species. The dragonfly Epitheca cynosura possesses fewer chromosomes (n = 10 or 11) than any other Epitheca species (n = 13m) but its estimated genome size is larger than E. canis (GS= 1.00 pg; n = 13m) and smaller than E. spinigera (GS= 1.32 pg; n = 13m). The lack of correspondence between changes in chromosome number and genome size in these exceptional cases may be caused by gross mutations such as chromosomal fusions and breaks that can result in the rearrangement of chromosomes with and without loss of DNA. 48 Flight strategy Odonate genome sizes showed a strong pattern in relation to flight strategy: fliers clearly have larger genomes than perchers (Fig. 2.2). This does not appear to be a taxonomic artefact (e.g., of dragonflies vs. damselflies), as the same pattern was found within dragonflies alone. This conclusion is reinforced by considerations of gliders, which fly actively without the benefit of highly efficient "flier" adaptations - and possess some of the smallest genomes among Odonata (Figure 2.2). Gliders may be described as specialized perchers that have evolved unique features and behaviours to be able to be on the wing most of the time. Corbet (1962) suggested that in order to accommodate their flight energy demands, gliders may have evolved specialized features to maximize food reserves (Corbet 1962). Despite their specialized features, gliders may have higher metabolic demands than any other percher, and this is reflected at the genome level with a stronger selection for smaller genomes. A second case is found in the minute damselflies Nehalennia Irene and N. integricolis. These species exhibit genomes that are considerably larger than any other perchers, and they do not fly in open areas but remain in small and highly populated areas (Forbes et al. 1997; Van Gossum et al. 2007). It may be suggested that the flight activity of these damselflies may be limited mainly to "hopping" between perches (usually sedge stems) (Fig. 2.2). This behaviour may have resulted in a reduction in flight metabolic rates, relaxation of the selective pressures exerted by flight metabolism on the genome, and an increase in genome size. Thus, "hoppers" and gliders may be considered opposite ends of a spectrum ranging from costly flight and small genome sizes to either very efficient or infrequent flight and large genomes. 49 Fliers showed the highest wing loading indexes as compared to perchers and fliers. May (1981) who obtained indexes from fresh odonates concluded also that wing loading index was higher in fliers than in perchers. However, the wing loading index in odonates is possibly not the main factor determining metabolic demands during flight. This is plausible since odonates can lift at least 2.5 times their weight (Marden 1987) and are often reported flying missing major portions of their wings. Thus, it is probably not wing size per se that determines flight energetics in dragonflies and damselflies, but some other parameter such as muscle activity or mass. Co-evolution of genome size and flight Perching is considered the ancestral flight strategy among dragonflies and damselflies, and remains the most common today. This strategy is associated with smaller genomes than seen among specialized fliers. It is therefore interesting to consider a historical link between change in flight strategy to efficient flying and an increase in genome size. Available phylogenetic hypotheses for odonates suggest that this association has evolved together at least once. As seen in Figure 2.3, there are two main scenarios, both of which suggest that low-cost flight and large genomes and/or high-cost "sprinting" (in perchers) or gliding and small genomes evolved together independently at least twice. In the first scenario, the families Gomphiidae and Libellulidae retained the percher strategy, while the flier strategy appeared independently in dragonfly families Aeshnidae, Cordulegastridae, and once before the split of the families Corduliidae and Macromiidae (Fig. 2.3A). The second (and more parsimonious) scenario involves the 50 independent evolution of a flier strategy once before the split of the flier families while the most recent lineage, the family Libellulidae, regained the ancestral, low-efficiency flight strategy (Fig. 2.3B). Genome size therefore would have both increased along with the appearance of the flier strategy and decreased independently in libellulids which reverted to a perching lifestyle. Genome size appears to have undergone increases in both situations, along with the decrease in energetic flight costs associated with higher flight efficiency. The patterns observed between flight and genome size seem to be reflected cytogenetically. Chromosome numbers are usually lower in percher dragonflies when compared to fliers. This may suggest that loss of DNA content is an ongoing process in odonates and that has resulted in the appearance of m-chromosomes (Oguma 1930). In addition, the loss of DNA may be discontinuous through fusions and breaks (; Dasgupta 1954; Cumming 1964; Kiauta 1967, 1968, 1969a, 169b) that resulted in gross losses of DNA and the origination of an autosomal remnant or m-chromosome. However, it appears that chromosome numbers have increased along with genome size independently in two odonate families. This may have occurred by continuous and small increases in copy number of transposable elements, by major duplications of long stretches of DNA, or duplication of an entire chromosome (aneuploidy). Genome size and flight Metabolic rates in relation to flight strategies have been found to be linked to genome size in bats and birds (Szarski 1970, 1983; Burton et al 1989; Tiersch and Wachtel 1991; Hughes and Hughes 1995, Waltari and Edwards 2002; Gregory 2002a, 2005). In birds, cell and genome sizes correlate positively and are usually smaller than 51 other vertebrates (Szarski 1970; Tiersch and Wachtel 1991; Gregory 2002a). Although avian genome sizes vary only two-fold, this was shown to correlate negatively with resting metabolic rate (Gregory 2002a). Additionally, genome size decreases in order from flightless birds (largest genomes), to weak, to moderate, to strong fliers (Hughes and Hughes 1995, Gregory 2002a, 2005). Recently, Organ et al. (2007) suggested that small genomes were an essential physiological adaptation that began in possibly endothermic theropod dinosaurs, before the evolution of birds and flight. Interestingly, odonates are the most similar among insects to flighted vertebrates in terms of flight mechanics and energetics. Odonates, unlike other insects, have synchronous direct flight in which flight muscles propel their wings directly, while wing beats are controlled by nerve pulses from motor neurons (Smith 1960, 1961a, b, 1966; Simmons 1977a, b; Josephson et al. 2000). The configuration of their flight muscle fibres resembles that of vertebrates (Smith 1961a, 1966). In addition, odonates, birds, and bats share very similar flying limitations in terms of maximum lift and wing beat frequencies (Marden 1987). The nature of the apparent constraints relating to flight probably differs between birds and odonates because insects do not possess respiratory blood cells. In fact, it does not seem that small genome size is an adaptation for flight per se in odonates, but rather that highly efficient flight allowed an increase in genome size in a subset of these insects. This contrasts with the situation in birds, where it is weak flyers and flightless groups that have the largest genomes. Nevertheless, the similarities are sufficient to indicate that flight imposes constraints on the evolution of genome size in both groups. These results 52 suggest that further examination of the association between physiology, ecology, and features at the genomic level is well worth pursuing in other insects. 53 Table 2.1. Odonata higher taxonomic classification including the number of genera and named species found in North America. This list is adapted from Westfall and May (1996) and Needham et al. (2000). Taxonomic updates were according to Schorr et al. (2008). Suborder Family # Genera # Named Species Zygoptera Calopterygidae 2 9 Coenagrionidae* 16 116 Lestidae 2 19 Megapodagrionidae 1 2 Plastystictidae 1 1 Protoneuridae 3 11 Pseudostigmatidae 2 2 Synlestidae 1 1 Anisoptera Aeshnidae 12 43 Cordulegastridae 1 8 Corduliidae 8 50 Gomphidae 14 109 Libellulidae 30 128 Macromiidae 2 10 Petaluridae 1 2 Total 96 511 The genera Argia (36) and Enallagma (37) contain the majority of species (73) belonging to this family. 54 Table 2.2. Published odonate genome sizes (GS, in pg) and means ± SE for major taxa. Taxonomic classification updated according to Schorr et al. (2008). Method abbreviations: RK) Reassociation kinetics; FD) Feulgen densitometry. References are provided at the end of the table. Taxonomic classification GS SE Method Reference Odonata 1.12 0.18 Zygoptera 1.10 0.40 Calopterygidae Calopteryx splendens 1.50 RK 1,2 55 Coenagrionidae Enallagma cyathigerum 0.70 RK Anisoptera 1.12 0.20 Aeshnidae 1.73 0.21 Aeshna juncea 1.00 - RK ' 1 Aeshna squamata 1.60 - RK 1,2 Rhinoaeschna bonariensis 1.81 0.17 FD 3 Rhinoaeschna confusa 2.16 0.16 FD 3 Taxonomic classification GS SE Method Reference Rhinoaeschna cornigera planaltica 2.08 0.08 FD Cordulegastridae Cordulegaster coronatus 1.70 RK Gomphidae 0.38 0.01 Lindenia tetraphylla coronatus 0.40 RK Gomphus flavipes 0.37 RK 56 Ophiogomphus cecilia 0.37 RK Libellulidae 0.66 0.01 Libellula quadrimaculata 0.67 RK Orthetrum ramburi 0.65 RK Sympetrum meridionale 0.65 RK References: 1) Petrov and Aleshin (1983); 2) Petrov et al. (1984); 3) Mola et al. (1994). Table 2.3. Hypothesized relationships between genome size and body size, developmental rate, flight strategy, and karyotype in Odonata. Parameter Hypothetical relationship with genome size Explanation Body size Positive: larger genomes associated with larger bodies. Larger genomes result in larger cells which result in larger bodies. Only applies when cell number remains relatively constant. Developmental rate Positive: longer life cycles are related to larger genomes. Larger genomes may increase the duration of the cell cycle, resulting in slower development. Flight strategy Negative: higher metabolic demands imposed by flight Oxygen uptake by cells is attained by diffusion from tracheal are associated with smaller genomes. tubes. This process may be optimized by having a small cell surface area to volume ratio as energetic demands increase. This results in smaller cells with smaller genomes. Karyotype Positive: smaller genomes are packaged into fewer Reduction of chromosome number in certain lineages may be chromosomes. due to loss of chromosomes (and DNA) rather than through fusion without DNA loss. Table 2.4. Classification of Odonata according to developmental rate (voltinism) (Corbet et al 2006). Voltinism mode Developmental rate Multivoltine > 3 generations per year Bivoltine 2 generations per year Univoltine 1 generation per year Semivoltine 1 generation every 2 years Partivoltine 1 generation in more than 2 years Table 2.5. Classification of Odonata according to flight strategies (Corbet 1962, 1996; May 1981). Flight Strategy Behaviour Groups Percher Remain on a perch most of the time and have short flight bursts. Zygoptera, Gomphiidae, and Libellulidae. Low efficiency (high cost) flying at high capacity ("sprinting"). Flier When active, remain on the wing most of the time. Aeshnidae, Cordulegastridae, Corduliidae, and High efficiency (low cost) flying below maximum capacity Macromiidae. ("long-distance walking"). Glider Fliers that are often migrant libellulids that make use of their A few libellulids (Trameinae) including: wide hindwings to glide and better sustain constant flight. Tramea Carolina, T. lacerata, and Myathiria Sustained flights without adaptations of other fliers. marcella. Table 2.6. Comparison between genome size estimates obtained from haemocytes and sperm for six individuals from four odonate species. Genome size estimates (in pg) for each tissue and from each individual are provided (paired t-test,^? = 0.32). The standard used for haemocytes estimates was haemocytes from Tenebrio molitor (GS = 0.52 pg; Juan and Petitpierre 1989). The standard used for spermatozoa was spermatozoa from Drosophila melanogaster Oregon R strain (GS = 0.18 pg; Rasch et al. 1971). In all cases at least 25 nuclei were measured. Species names follow Schorr et al. (2008). Species GS Haemocytes GS Sperm Enallagma hageni 1 1.12 1.25 60 Enallagma hageni 2 1.18 1.25 Ophiogomphus rupinsulensis 0.89 0.94 Lestes rectangularis 0.7,7 0.73 Hetaerina americana 1 1.11 Hetaerina americana 2 _ 1.08* * Only 13 nuclei could be measured due to poor sample quality. Table 2.7. Odonate genome size estimates (GS, in pg), standard error (± SE), and number of individuals (N) for 62 dragonfly species and 38 damselfly species. Collection site and year (L) are included for each species. Haploid chromosome number (K) and source reference (R) are also included. Species names and taxonomic classification follow Schorr et al. (2008). All abbreviations used are provided at the end of the table. Taxonomy GSJpg) + /-SE N K/R Suborder Anisoptera (Dragonflies) 1.01 0.05 Family Aeshnidae 1.61 0,10 Genus Aeshna 61 Aeshna canadensis 2.20 — 1 14/A 3 Aeshna constricta 1.76 0.06 4 1 Aeshna eremita 1.85 - 1 12 Aeshna tuberculifera 1.78 0.10 2 3,12 Aeshna umbrosa 2.00 - 1 14/A 2 Aeshna verticalis 1.59 1 11 Genus Anax Anax Junius 1.44 14, 14m/A Genus Basiaeschna Basiaeschna janata 1.16 13/A 14 Taxonomy GS(pg) + /-SE N K/R Genus Epiaeschna Epiaeschna heros 1.44 17 Genus Gomphaeschna Gomphaeschnafurcillata 1.20 12 Genus Nasiaeschna Nasiaeschna pentacantha 1.31 17 Family Corduleeastridae 0.94 Genus Cordulegaster Cordulegaster maculata 0.94* 13m/A 12 Family Corduliidae 1.35 0.16 Genus Cordulia Cordulia shurtleffi 1.54* 13/A 11 Genus Dorocordulia Dorocordulia libera 0.98 0.03 6, 7/A 12 Genus Epitheca Epitheca canis 1.00 0.02 3 13m/A 17 Epitheca cynosura 1.10 0.06 4 10, 11 /A 17 Epitheca princeps 0.98* 1 17 Taxonomy GS(pg) + /-SE N K/R Epitheca spinigera 1.32 13m/A 12 Genus Neurocordulia Neurocordulia yamaskanensis 1.08 12 Genus Somatochlora Somatochlora williamsoni 1.80 12 Somatochlora elongata 2.39* Family Gomphiidae 0.76 0.03 Genus Arigomphus Arigomphus pallidus 0.83 12/B 17 63 Arigomphus villosipes 0.82 0.03 1 Genus Dromogomphus Dromogomphus spinosus 0.57 12m/A 14 Genus Gomphus Gomphus cavillaris 0.71 1 - 17 Gomphus (Gomphurus) dilatatus 0.76 0.05 2 - 17 Gomphus exilis 0.71 0.03 4 12m, 12 / A, C 11,12,14 Gomphus (Hylogomphus) geminatus 0.78 1 - 17 Gomphus graslinellus 0.73 1 11 Taxonomy GS(pg) + /-SE N K/R Gomphus minutus 0.75 — 17 Gomphus spicatus 0.72 - 12*7 A 11 Genus Hagenius Hagenius brevistylus 0.93 - Genus Ophiogomphus Ophiogomphus rupinsulensis 0.94 0.04 12m/A 12 Genus Progomphus Progomphus obscurus 0.61 0.04 12m/A 17 Genus Stylogomphus 64 Stylogomphus albistylus 0.72 0.04 12 Family Libellulidae 0.77 0.03 Genus Brachymesia Brachymesia gravida 0.69 - 17 Genus Celithemis Celithemis bertha 0.87 - 1 - 17 Celithemis elisa 0.77 0.06 2 13m/A 11 Celithemis eponina 0.91 - 1 - 17 Celithemis ornata 0.54 _ 1 13m/D 17 Taxonomy GS(pg) + /-SE N K/R Genus Erythemis Erythemis simplicicollis 0.56 13ra/A 13 Genus Erythrodiplax Erythrodiplax minnscula 0.67 13m/D 17 Genus Ladona Ladona julia 0.62 0.03 3 13m/A .11,14 Ladona deplanata 0.60 0.02 2 - 17 Genus Leucorrhinia Leucorrhinia glaciallis 0.98 — 1 13m/A 15 65 Leucorrhinia hudsonica 0.94 - 1 13m, 13/A 12 Leucorrhinia intacta 0.93 - 1 13m, 13/A 11 Leucorrhinia proxima 1.27 - 1 13m/A 16 Genus Libellula Libellula incesta 0.74 — 1 13/A 14 Libellula luctuosa 0.87 0.04 2 13/E 4,5 Libellula pulchella 0.84 0.03 3 13m/A 2 Libellula vibrans 0.95 1 13m7A 17 Taxonomy GS(pg) + /-SE N K/R L Genus Miathyria Miathyria marcella 0.41 0.008 2 13, 13m/B,G 18 Genus Perithemis Perithemis tenera 0.68 0.009 3 13m/B,D 13 Genus Plathemis Plaihemis lydia 0.62 0.05 2 13m/A .2 Genus Sympetrum Sympetrum internum 0.78 0.07 2 13m/A 1,6 Sympetrum obtrusum 0.82 0.03 3 m 66 13 /A 1,6 Sympetrum vicinum 0.77 0.006 3 5 Genus Tramea Tramea Carolina 0.67 1 13/A 17'~ Tramea lacerata 0.67* 1 13/A 1 Family Macromiidae 1.08 <0.01 2 Genus Didymops Didymops transversa 1.08* 1 13m/A 17 Genus Macromiia Macromiia illionensis georgina 1.07* 1 17 Taxonomy GS(pg) + /-SE N K/R Suborder Zygoptera (Damselflies) 1.02 0.04 38 Family Calopterygidae 1.03 0.04 4 Genus Calopteryx Calopteryx maculata 1.00 0.03 7 13m/A,B 1,7, Calopteryx aequabilis 1.11 0.02 4 13m/A 1,7 Calopteryx dimidiata 0.94 1 13m/D 17 Genus Hetaerina Hetaerina americana 1.11* 13m/A,B Family Coenaarionidae 1.11 0.04 6/ Genus Amphiagrion Amphiagrion saucium 0.89 12 Genus Argia Argia apicalis 0.88 i 7 Argia bipuctulata 0.94 0.05 2 17 Argia fumipennis 0.93 0.02 11 14/D 1,18 Argia moesta 0.90 0.04 5 13, 14,17 Argia sedulla 0.90 0.02 3 14/B 17 Argia tibialis 0.88 1 17 Taxonomy GS(pg) + /-SE N K/R L Genus Enallagma Enallagma annexum 1.14 - 1 - 12 Enallagma antennatum 1.35 - 1 - 1,2 Enallagma basidens 0.94 - 1 - 13 Enallagma boreale 1.26 0.006 2 14/A 15 Enallagma carunculatum 1.00 - 1 14/A 7 Enallagma civile 1.10 0.05 6 14/A 1,9 Enallagma cyathigerum 1.20 0.02 3 14, 15/A,C,F 1 Enallagma doubledayi 1.15 0.03 6 - 17,18 68 Enallagma durum 1.20 - 1 - 17 Enallagma ebrium 1.18 0.02 12 14/A 1,2; 3 Enallagma exsulans 1.18 0.03 9 - i;7 Enallagma geminatum 1.08 - 1 - 12 Enallagma hageni 1.19 0.03 4 - 11,14 Enallagma signatum 1.28 - 1 - 1 Enallagma vesperum 1.34 1 _ 14 Genus Ischnura Ischnura posita 0.96 0.04 Taxonomy GS (pg) +/-SE N K/R L Ischnura ramburii 0.88 0.02 3 14m/D 17,18 Ischnura verticalis 0.97 0.03 5 14/A 1 Genus Nehalennia Nehalennia integricollis 1.53 0.04 7 - 17 Nehalennia irene 1.80 0.04 3 14/A 13,14 Family Lestidae 0.64 .0.02 Genus Lestes Lestes congener 0.60 0.03 4 13m/A 1,13 Lestes dryas 0.72 0.02 7 13, 13m7A 1,2,12 69 Lestes eurinus 0.60 - 1 - 12 Lestes forcipatus 0.63 0.01 3 11/A 13 Lestes inaequalis 0.59 - 1 - 12 Lestes rectangularis 0.73 0.04 4 •13m/A 2,13,9 Lestes unguiculatus 0.62 0.04 4 - .2,10 Genome size estimates were obtained using FLA and by measuring DNA content in at least 25 spermatozoa nuclei in every individual. The standard used to calculate estimates was spermatozoa nuclei from the fruit fly Drosophila melanogaster Oregon R strain (GS = 0.18 pg). * Estimates based on haemocyte samples using haemocytes from Tenebrio molitor (GS = 0.52 pg) as the standard. Karyotype (K) references: A) Cruden (1968), B) Cumming (1964), C) Kiauta (1969a), D) Kiauta and van Brink (1978), E) Smith (1916), F) van Brink and Kiauta (1968), G) Ferreira et al. (1979). Collection information including location (L), season, and year: 1) Guelph (ON), summer (06); 2) Ariss (ON), summer (06); 3) Haliburton (ON), summer (06); 4) Cambridge (ON), summer (06); 5) Muskrat Lake (ON), summer (06); 6) Kashagawigamog Lake (ON), summer (06); 7) Thames river, Thamesville (ON), summer (06); 8) Crowe river, Peterborough (ON), summer (06); 9) Terra cotta (ON), summer (06); 10) Katchawanooka Lake (ON), summer 2006; 11) Guelph (ON), summer (07); 12) Algonquin Park (ON) / summer (07); 13) Royal Botanical Gardens, Hamilton (ON), summer (07); 14) Fletcher Lake (ON), summer (07); 15) Livingston Lake (ON), summer (07); 16) Wollaston Lake (ON), summer (07); 17) Tallahassee (FL), May (07); 18) Archbold Research Station, Lake Placid (FL), May (07). Table 2.8. Odonate body size measurements, including means and standard error, for each species for minimum body weight (BDW, in grams) and total length (in mm) for body (BL), head (HL), thorax (THL), abdomen (AL), forewing (FWL), and hindwing (HWL). Also included are areas (in mm2) for forewing (FWA), and hindwing (HWA). Species names and taxonomic classification were updated according to Schorr et al. (2008). Taxonomy N BDW BL HL THL AL FWL HWL FWA HWA Suborder Anisoptera (dragonflies) Family Aeshnidae Genus Aeshna Aeshna canadensis 0.159 67.58 6.83 11.81 48.94 45.47 45.40 369.57 483.61 71 (Standard error) (0.028) (1-67) (0.38) (0.92) (2.96) (0.17) (0.58) (8.32) (4.32) Aeshna constricta 0.228 71.97 6.84 10.72 54.41 46.00 45.48 411.15 517.38 (0.022) (1.41) (0.23) (0.39) (1.26) (0.57) (0.55) (9.79) (14.10) Aeshna eremita 0.312 78:80 6.27 16.19 56.33 51.24 49.47 460.72 574.33 Aeshna tuberculifera 0.218 72.44 6.62 12.53 53.29 49.84 48.82 468.74 591.42 (0.017) (1.88) (0.23) (1.25) (2.91) (1.01) (1.41) (23.90) (28.94) Taxonomy N BDW BL HL THL AL FWL HWL FWA HWA Aeshna umbrosa 1 0.207 73.05 4.92 11.11 57.03 46.51 45.97 382.48 481.44 Aeshna verticalis 1 0.220 66.25 6.48 11.64 48.13 45.91 44.91 400.40 486.97 Genus Anax Anaxjunius 4 0.344 73.91 7.53 14.24 52.14 50.45 49.63 469.87 598.43 Genus Basiaeschna (0.060) (2.13) (0.43) (0.91) (2.66) (1.08) (0.99) (19.71) (30.81) Basiaeschna janata 1 38.79 37.35 35.90 224.23 258.94 Genus Epiaeschna 0.128 55.09 5.74 10.56 Epiaeschna hews 1 61.35 55.99 54.59 527.92 632.77 Genus Gomphaeschna 0.392 82.39 8.05 12.99 Gomphaeschna furcillata 1 37.82 33.55 32.02 194.50 226.06 Genus Nasiaeschna 0.029 52.22 5.28 9.12 Nasiaeschna pentacantha 1 0.263 73.61. 6.62 12.22 54.77 50.18 50.17 447.13 511.97 Family Cordulegastridae Genus Cordulegaster Taxonomy N BDW BL HL THL AL FWL HWL FWA HWA Cordulegaster maculata 1 0.944 0.19 68.30 4.99 13:75 49.55 41.63 39.96 277.29 Family Corduliidae Cordulia shurtleffi 0.070 43.68 5.21 9.89 28.58 29.52 28.11 159.05 188.18 Genus Dorocordulia Dorocordulia libera 4 0.046 38.46 3.62 7.59 27.25 27.98 27.43 152.04 193.53 (0.001) (0.78) (0.44) (0.24) (0.19) (0.21) (0.29) (4.21) (5.64) Genus Epitheca 73 Epitheca canis 2 0.068 42.44 4.46 9.32 28.66 29.43 28.58 169.09 210.37 (0.006) (1.36) (0.60) (0.56) (0.20) (0.83) (0.72) (0.67) (4.09) Epitheca cynosura 4 0.055 38.28 4.95 7.69 25.64 28.27 27.17 159.53 197.10 (0.002) (0.83) (0.17) (0.15) (0.82) (0.44) (0.54) (6.19) (6.67) Epitheca princeps 1 0.155 57.94 5.10 12.14 40.70 43.87 42.22 329.21 409.38 Epitheca spinigera 1 0.071 41.19 5.08 8.41 27.70 30.47 29.73 171.35 225.59 Genus Neurocordulia Taxonomy N BDW BL HL THL AL FWL HWL FWA HWA 1 0.121 52.49 4.93 9.38 38.18 33.74 32.30 214.00 256.25 Neurocordulia yamaskanensis Genus Somatochlora 1 0.106 53.34 5.59 9.30 38.44 38.76 38.44 259.24 304.73 Somatochlora elongata 1 0.120 57.43 5.18 9.22 43.03 38.23 37.54 264.96 321.12 Somatochlora williamsoni Family Gomphiidae Genus Arigomphus 1 0.129 56.30 4.32 10.74 41.24 34.60 33.07 204.21 246.25 . Arigomphus pallidus Arigomphus villosipes 2 0.101 49.78 - 11.00 38.79 32.87 31.51 184.69 215.01 (0.004) (0.30) (0.32) (0.02) (0.36) (0.63) (4.72) (2.77) Genus Dromogomphus Dromogomphus spinosus 2 0.117 55.75 3.83 11.77 40.16 34.86 34.03 224.35 274.55 (0.007) (0.55) (0.07) (0.24) (0.86) (0.63) (0.21) (9.90) (3.08) Genus Gomphus Gomphus cavillaris 1 0.052 43.34 3.83 9.62 29.89 27.39 26.29 118.13 150.96 Taxonomy N BDW BL HL THL AL FWL HWL FWA HWA Gomphus (Gomphurus) dilatatus 3 0.220 73.10 4.74 15.43 52.93 41.01 39.11 296.84 350.09 (0.031) (0.82) (0.16) (0.42) (0.36) (0.55) (0.54) (2.84) (4.35) Gomphus minutus 1 0.074 47.21 2.99 10.67 33.55 27.50 27.75 154.65 171.99 Gomphus exilis 5 0.054 41.62 3.27 8.93 29.43 25.26 23.72 121.78 139.04 (0.001) (0.94) (0.19) (0.16) (0.75) (0.51) (0.53) (3.67) (3.32) Gomphus (Hylogomphus) geminatus 1 0.102 53.25 4.38 10.74 38.13 33.22 31.47 182.20 231.28 Gomphus graslinellus 1 0.096 47.04 4.65 9.36 33.03 31.09 29.39 183.97 211.76 75 Gomphus exilis 5 0.054 41.62 3.27 8.93 29.43 25.26 23.72 121.78 139.04 Gomphus minutus 1 0.074 47.21 2.99 10.67 33.55 27.50 27.75 154.65 171.99 Gomphus spicatus 1 0.064 45.17 4.10 8.47 32.60 25.66 . 24.58 124.61 144.33 Genus Hagenius 0.317 73.08 5.03 15.29 52.75 48.89 46.19 409.66 478.08 Hagenius brevistylus 4 (0.012) (0.79) (0.15) (0.30) (0.845) (0.25) (0.21) (9.39) (6.60) Genus Ophiogomphus Taxonomy N BDW BL HL THL AL FWL HWL FWA HWA Ophiogomphus rupinsulensis 2 0.103 49.38 3.86 10.72 34.80 30.16 28.94 177.64 204.26 (0.003) (0.46) (0.01) (0.10) (0.36) (0.02) (0.41) (3.88) (5.56) Genus Progomphus Progomphus obscurus 7 0.078 52.82 4.22 10.49 38.11 32.90 31.27 202.79 237.00 (0.004) (0.60) (0.09) (0.19) (0.53) (0.63) (0.51) (6.28) (5.92) Genus Stylogomphus Stylogomphus albistylus 2 0.068 35.90 3.14 6.76 26.00 22.63 21.33 99.64 117.75 76 (0.004) (1.61) (0.60) (0.35) (0.66) (0.37) (0.79) (7.19) (10.76) Family Macromiidae Genus Dydimops Didymops transversa 1 0.134 53.11 5.48 11.61 36.01 37.63 35.89 245.24 294.73 Genus Macromia Macromia illionensis georgina 1 0.214 70.15 6.14 13.67 50.34 46.32 44.25 366.67 392.09 Family Libellulidae Genus Brachymesia Taxonomy N BDW BL HL THL AL FWL HWL FWA HWA 1 0.057 45,80 • 4.34 9.16 32.30 38.94 35.35 234.59 268.40 Brachymesia gravida Genus Celithemis 2 0.032 31.09 3.88 7.15 20.06 26.22 24.25 134.70 158.73 Celithemis bertha (0.002) (0.74) (0.15) (0.29) (1.18) (0.08) (0.02) (4.27) (0.06) Celithemis elisa 2 0.031 30.97 3.92 7.18 19.86 27.03 26.16 131.70 178.80 (0.001) (0.57) (0.19) (0.04) (0.41) (1.09) (0.87) (10.08) (11.83) Genus Erythemis Erythemis simplicicollis 2 0.053 42.58 4.02 10.86 27.70 32.54 31.56 186.03 225.79 (0.019) (0.65) (0.05) (0.31) (0.92) (0.31) (0.22) (4.89) (1.43) Genus Erythrodiplax Erythrodiplax minuscula 1 0.019 25.30 2.63 6.23 16.44 20.76 19.98 94.35 113.73 Genus Ladona Ladona deplanata 2 0.044 34.25 4.09 8.31 21.85 27.55 26.01 148.92 178.88 (0.004) (0.88) (0.07) (0.43) (0.52) (0.466) (0.88) (10.60) (17.23) Taxonomy N BDW BL HL THL AL FWL HWL FWA HWA Ladona julia 3 0.084 39.87 4.16 10.85 24.86 34.81 33.16 210.17 258.81 (0.004) (0.98) (0.15) (0.38) (0.92) (1.00) (0.91) (6.60) (8.53) Genus Leucorrhinia Leucorrhinia proxima 3 0.040 35.13 3.75 7.58 23.80 26.72 25.77 142.72 175.25 (0.004) (0.20) (0.02) (0.36) (0.31) (0.94) (1.25) (4.09) (8.36) Leucorrhinia intacta 2 0.041 32.69 3.94 9.48 19.27 25.63 24.58 131.92 163.20 (0.001) (0.34) (0.06) (0.39) (0.04) (0.94) (1.13) (4.91) (9.00) 78 1 0.024 29.65 3.26 6.95 19.44 23.68 22.94 113.75 139.07 Leucorrhinia hudsonica 1 0.040 34.30 4.34 7.49 22.48 27.17 25.92 142.98 173.73 Leucorrhinia glacialis Genus Libellula 2 0.088 42.59 4.37 10.02 28.20 36.74 35.54 254.00 329.27 Libellula luctuosa (0.005) (0.03) (0.24) (0.49) (0.28) (0.48) (0.35) (10.81) (9.97) ' Libellula pulchella 3 0.151 50.10 5.32 12.76 32.02 41.40 39.57 312.59 373.56 (0.012) (2.14) (0.39) (0.67) (1.38) (0.54) (0.78) (17.71) (19.79) Taxonomy N BDW BL HL THL AL FWL HWL FWA HWA Libellula incesta 1 0.112 47.86 4.77 12.52 30.58 40.67 39.21 266.31 330.14 Libellula vibrans 4 0.210 60.01 5.27 14.06 40.68 49.31 46.44 419.17 506.93 (0.007) (1.56) (0.30) (0.52) (0.99) (1.18) (0.97) (14.07) (19.57) Genus Miathyria Miathyria marcella 5 0.029 39.35 4.74 8.05 26.56 34.80 33.09 222.08 293.77 (0.001) (0.43) (0.15) (0.23) (0.42) (0.42) (0.38) (7.79) (8.42) Genus Perithemis 79 Perithemis tenera 3 0.027 23.63 3.41 6.94 13.29 19.41 18.94 86.47 106.70 (0.002) (0.61) (0.35) (0.36) (0.52) (0.10) (0.21) (1.570) (5.07) Genus Plathemis Plathemis lydia 1 0.111 44.13 4.47 12.41 27.25 31.73 30.80 202.24 242.70 Genus Sympetrum Sympetrum internum 2 0.026 31.47 3.67 6.72 21.09 25.36 24.28 136.04 167.01 (0.003) (0.32) (0.29) (0.14) (0.18) (0.14) (0.04) (3.73) (6.63) Taxonomy N BDW BL HL THL AL FWL HWL FWA HWA 12 0.022 32.57 3.63 6.71 22.22 24.20 23.23 127.57 155.51 Sympetrum obtrusum (0.001) (0.46) (0.14) (0.17) (0.49) (0.31) (0.28) (3.18) (3.84) 4 0.022 32.15 3.35 7.17 21.63 23.95 22.88 119.23 147.94 Sympetrum vicinum (0.003) (0.32) (0.29) (0.14) (0.18) (0.14) (0.04) (3.73) (6.63) Genus Tramea 1 0.133 48.96 5.02 9.44 34.49 45.34 43.27 361.02 519.67 1 0.061 50.29 6.05 10.15 34.09 45.04 43.19 367.51 506.57 Tramea Carolina Tramea lacerata Suborder Zygoptera (damselflies) Family Calopterygidae 6 0.038 48.41 2.53 7.34 38.55 32.57 31.19 202.74 202.47 Genus Calopteryx (0.001) (0.88) (0.07) (0.16) (0.88) (0.36) (0.44) (5.64) (7.04) Calopterix aequabilis Calopteryx dimidiata 3 0.026 48.94 2.83 7.12 38.99 31.69 30.85 199.93 186.19 (0.001) (0.88) (0.07) (0.16) (0.88) (0.36) (0.44) (5.64) (7.04) Taxonomy N BDW BL HL THL AL FWL HWL FWA HWA Calopteryx maculata 3 0.024 43.79 2.46 6.20 35.13 29.27 28.34 199.60 201.57 (0.005) (0.90) (0.21) (0.75) (1.46) (1.10) (1.03) (9.31) (12.58) Genus Hetaerina Hetaerina americana 8 0.024 45.03 2.40 6.67 35.96 27.54 26.43 122.21 108.82 (0.001) (0.56) (0.06) (0.13) (0.53) (0.44) (0.36) (3.67) (3.90) Family Coenagrionidae Genus Amphiagrion Amphiagrion saucium 1 0.004 23.67 1.39 3.54 18.74 - - Genus Argia Argia apicalis 3 0.011 36.47 1.94 5.53 29.00 22.44 20.99 74.18 65.11 (0.001) (0.33) (0.07) (0.25) (0.33) (0.39) (0.33) (1.53) (0.09) Argia bipunctulata 3 0.006 28.63 1.78 4.13 22.72 16.54 15.81 44.45 41.05 (0.001) (0.06) (0.10) (0.37) (0.27) (0.20) (0.19) (0.82) (0.18) Argia fumipennis 13 0.008 31.64 1.67 4.84 25.13 19.47 19.11 60.63 57.00 (0.001) (0.31) (0.05) (0.11) (0.23) (0.27) (0.20) (2.31) (1.17) Taxonomy N BDW BL HL THL AL FWL HWL FWA HWA Argia moesta 12 0.014 40.89 2.16 5.97 32.77 25.62 23.99 85.43 77.17 (0.001) (0.57) (0.05) (0.08) (0.54) (0.25) (0.25) (1.69) (1.88) Argia sedulla 4 0.008 33.50 1.88 5.03 26.59 19.97 19.09 60.23 55.58 (0.001) (0.29) (0.12) (0.06) (0.28) (0.17) (0.14) (0.89) (0.55) Argia tibialis 3 0.009 34.07 1.83 4.93 27.31 20.76 20.10 66.95 61.67 (0.001) (0.68) (0.06) ' (0.17) (0.62) (0.48) (0.50) (0.62) (0.25) Genus Enallagma 82 1 0.012 35.37 1.90 5.70 27.77 21.15 19.86 54.65 50.03 Enallagma annexum Enallagma antennatum 2 0.008 29.14 1.75 4.56 22.83 , 18.33 17.32 42.21 39.65 (0.001) (2.13) (0.06) (0.11) (1.56) (0.25) (0.26) (0.72) (0.63) Enallagma basidens 1 0.005 26.18 1.26 4.07 20.85 - - - - Enallagma boreale 2 0.009 30.57- 1.91 4.85 23.81 19.31 18.09 49.00 41.66 (0.001) (0.43) (0.02) (0.16) (0.29) (0.21) (0.27) (3.75) (3.81) Enallagma carunculatum 1 0.012 34.85 1.85 5.69 27.31 19.78 17.87 54.31 47.02 Taxonomy N BDW BL HL THL AL FWL HWL FWA HWA Enallagma civile 6 0.010 32.44 1.95 5.02 25.47 19.02 17.82 50.82 45.27 (0.001) (1.22) (0.04) (0.21) (1.10) (0.46) (0.41) (2.44) (1.87) Enallagma cyathigerum 1 0.012 31.51 1.78 5.11 24.62 19.80 18.36 56.69 50.62 Enallagma doubledayi 6 0.007 31.14 1.75 4.65 24.74 17.92 16.84 43.32 38.28 (0.001) (0.-38) (0.11) (0.15) (0.25) (0.35) (0.24) (0.10) (1.02) Enallagma durum 1 0.017 39.77 2.32 6.27 31.19 22.88 21.34 39.67 39.77 Enallagma ebrium 13 0.008 28.96 1.69 4.64 22.63 17.74 16.43 41.15 35.92 (0.001) (0.38) (0.03) (0.11) (0.33) (0.15) (0.15) (0.65) (0.61) Enallagma exsulans 11 0.007 32.49 1.59 4.54 26.36 19.31 18.08 45.30 40.71 (0.001) (0.48) (0.06) (0.08) (0.46) (0.25) (0.24) (1.04) (0.94) Enallagma geminatum 2 0.007 25.15 1.57 4.28 19.30 15.83 14.63 31.15 26.87 (0.002) (0.17) (1.46) (0.10) (0.42) (0.59) (0.57) (1.87) (1.26) Enallagma hageni 4 0.006 29.30 1.69 4.41 23.20 17.81 16.35 38.86 33.99 (0.001) (0.86) (0.13) (0.10) (0.71) (0.27) (0.25) (1.15) (1.30) Taxonomy N BDW BL HL THL AL FWL HWL FWA HWA Enallagma signatum 2 0.006 31.44 1.32 4.56 25.55 17.72 16.80 42.07 37.56 (0.001) (1.21) (0.01) (0.06) (1.15) (0.21) (0.06) (0.28) (0.55) Enallagma vesperum 1 0.006 32.32 1.40 4.91 26.01 18.27 16.62 31.96 32.32 Genus Ischnura Ischnura posita 6 0.004 25.72 1.35 3.65 20.72 13.35 12.56 22.78 20.96 (0.001) (0.62) (0.06) (0.21) (0.50) (0.42) (0.36) (0.129) (1.08) Ischnura ramburii 4 0.008 31.32 1.60 5.07 24.64 16.91 15.82 38.79 34.96 (0.001) (0.65) (0.10) (0.15) (0.42) (0.59) (0.57) (3.06) (2.83) Ischnura verticalis 5 0.005 26.93 1.50 4.38 21.05 13.56 12.88 24.97 22.18 (0.001) (0.63) (0.08) (0.10) (0.57) (0.29) (0.34) (0.92) (1.037) Genus Nehalennia Nehalennia integricollis 4 0.002 22.82 1.11 3.15 18.57 12.07 11.45 21.04 19.39 (0.001) (0.21) (0.03) (0.10) (0.20) (0.15) (0.17) (0.27) (0.38) Taxonomy N BDW BL HL THL AL FWL HWL FWA HWA Nehalennia irene 4 0.004 26.92 1.32 3.94 21.66 15.21 14.41 32.92 30.22 (0.001) (0.48) (0.08) (0.08) (0.54) (0.60) (0.55) (2.92) (2.45) Family Lestidae Genus Lestes Lestes congener 4 0.012 34.38 1.66 5.37 27.34 19.83 18.89 55.32 50.89 (0.001) (0.54) (0.09) (0.16) (0.61) (0.19) (0.14) (1.35) (1.36) Lestes dryas 7 0.017 36.17 2.22 5.83 28.12 20.87 20.12 64.54 59.32 (0.001) (0.51) (0.05) (0.11) (0.50) (0.27) (0.22) (0.93) (0.84) 85 Lestes eurinus 1 0.029 46.01 2.50 7.47 36.03 28.75 27.74 110.54 96.50 Lestes forcipatus 3 0.015 35.74 2.11 6.02 27.61 21.18 20.28 60.97 54.41 (0.001) (0.51) (0.11) (0.16) (0.26) (0.36) (0.21) (2.05) (0.85) Lestes inaequalis 1 0.022 43.79 2.40 7.15 34.23 25.81 24.74 80.73 71.67 Lestes rectangularis 5 0.013 44.64 1.91 6.25 36.48 21.55 21.10 65.22 59.75 (0.002) (1.05) (0.07) (0.08) (1.18) (0.61) (0.37) (3.30) (2.54) Lestes unguiculatus 3 0.015 35.80 2.18 5.63 27.99 19.67 18.75 63.60 55.62 (0.001) (0.72) (0.20) (0.09) (0.43) (0.18) (0.08) (2.95) (2.98) A B mm mm jMHpWBiuui'i" i' te?° 'Wii^^%!*%.™ ™ ' 86 Figure 2.1. Body size and wing area measurements of a representative odonate (the damselfly Enallagma erbium), using NIS-Elements BR Software. Lengths for the head, thorax, and abdomen were measured separately and summed to obtain total body length (A). Measurements also included lengths (A) and areas (B) for one forewing and one hind wing. All measurements were obtained in pixels and converted to millimetres using Microsoft Excel (2003-2007). A 10 mm2 section of millimetre-square graph paper was included in each picture and used as a scale. 2.5 C | 2.0 - • "3 1-5 ~ b I • a. (2) w ° 1.0- 1 ! 1 •a (23) 0.5 ~ 1 1 (36) t36' (3 u .Gliders Percher Percher Hoppers Fliers 87 Dragonflies Damselflies Flight Strategy Figure 2.2 Genome size (GS) and flight strategies in Odonata. Genome size ranges (black bars) and means (white lines), and number of species are presented for each flight strategy. Different letters between each column represent significant differences (p < 0.05) between means. All damselfly species are perchers except for Nehalennia integricolis (GS = 1.80 pg) and N. Irene (GS = 1.53 pg) which are shown as "hoppers" since they remain in small areas as adults and their "flight" may be limited to "jumping" between branches. Percher dragonflies are members of the families Gomphiidae and Libellulidae. Gliders were the libellulids Miathyria marcella (GS = 0.41 pg), Tramea lacerata (GS = 0.67 pg), and T. Carolina (GS = 0.67 pg). Fliers are members of the dragonfly families Aeshnidae, Cordulegastridae, Cordullidae, and Macromiidae (Corbet 1960, 1996; May 1981). Hypothesis 1 Hypothesis 2 A B Gomphiidae 0.76pg ± 0.03 (0.54-0.94, n = Gomphiidae" t Aeshnidae 1.61 pg± 0.03 (1.16-2.20, iv Aeshnidae * t Cordulegastridae 0.94 (n= 13" Cordulegastridae • t — "Corduliidae 1.35pg± 0.16 (0.98-2.36, rv Corduliidae * t^iacromiidae 1.08pg±0.01 (1.07-1.08, rv Macromiidae I 'Libellulidae 0.77pg ±0.03 (0.41-1.18, rv Libellulidae* Figure 2.3. Evolution of genome size, flight strategy, and karyotypes in dragonflies. Branches representing fliers are denoted in grey and perchers in black. The flier strategy is associated with large genomes while the percher strategy is associated with small genomes. A) Hypothesis 1: Fliers evolved independently from an ancestral percher which was retained in the families Gomphiidae and Libellulidae whereas the other families each independently evolved flier lifestyles. B) Hypothesis 2: The flier strategy evolved once before the split of the families Aeshnidae, Cordulegastridae, Corduliidae and Macromiidae while the family Libellulidae regained the ancestral percher lifestyle independently. The high activity but metabolically efficient flier strategy is associated with an increase in genome size (arrow up) while the percher strategy is associated with a decrease (arrow down). Means and standard error, and ranges for genome size (GS) and the karyotype (n) for each family are shown. The presence of an m-chromosome is denoted with "m." Phylogeny adapted from Misof et al. (2003) and Ware et al. (2007). Chapter 3: Genome size diversity, parasitism, and eusociality in Hymenoptera: a first approach to the superfamilies Apoidea, Chalcidoidea, Ichneumonoidea, and Vespoidea 89 ABSTRACT Estimates for 87 hymenopteran species from the superfamilies Apoidea, Chalcidoidea, Ichneumonoidea, and Vespoidea were used to examine for the first time the genome size variation for the order Hymenoptera and possible relationships with two of its most characteristic traits: parasitism and eusociality. Based on this study, genome size did not correlate with parasitism or eusociality. However, the order Hymenoptera appears to exhibit the smallest genome size range and the lowest maximum and mean values among holometabolous insect orders studied thus far. The smallest genome sizes for the order were found in the parasitoid family Braconidae; however, estimates for parasitoid wasps from the superfamily Chalcidodea did not differ significantly from non- parasitoid groups examined in this study. Similarly, there was no significant difference in genome size estimates between solitary and eusocial groups when examined in the superfamilies Vespoidea and Apoidea. 90 Introduction The order Hymenoptera is one of the most species-rich animal groups on the planet (125,000 named species), best known for their array of life strategies (e.g. parasitism and eusociality) and for having a haploid/diploid sex determination system (Grimaldi and Engel 2005). Hymenopterans are usually classified into two suborders: the primitive (but paraphyletic) Symphyta (sawflies and wood wasps) and the Apocrita ("true" wasps) (Grimaldi and Engel 2005). The Symphyta as Grimaldi and Engel (2005) stated "are, simply put, anything that is not apocritan." In addition, the composition and phylogenetic relationships between and within the 14 superfamilies recognized for this "suborder" remain unclear. The Apocrita is subdivided into 14 superfamilies containing most of the species diversity for the order (roughly 100,000 species), but the phylogenetic relationships between them are still uncertain and are presented here as summarized by Grimaldi and Engel (2005) (Figure 3.1). However, there is general agreement in the grouping of the "Aculeata" (superfamilies Apoidea, Chrysidoidea, and Vespoidea) and the phylogenetic relationships between and within these groups at the family level (Figure 3.2) (Brothers and Carpenter 1993; Brothers 1999; Grimaldi and Engel 2005). The hymenopterans that have attracted the most attention for their economic importance and unique characteristics are those belonging to the superfamilies Apoidea, Chalcidoidea, Ichneumonoidea, and Vespoidea. The first two include some of the most commonly used biological control agents such as the parasitoids Nasonia vitripennis (Chalcidoidae) and Habrobracon juglandis (Ichneumonoidae), while the remaining 91 groups include the most important pollinators (bees), some of the most abundant and familiar insects (e.g. ants), and the only eusocial groups found in the order which include ants, bees, and some of the stinging wasps from the family Vespidae (Holldobler and Wilson 1990, Michener 2000; Richter 2000). Hymenoptera genomics Traditionally, hymenopterans are best known for having a haploid/diploid sex determination system in which females and males develop from fertilized and unfertilized eggs, respectively. In some cases, sex is determined by a single locus known as complementary sex determination (CSD) where heterozygous individuals develop as females while homozygous or hemizygous individuals with a single copy develop as males (Page et.al. 2002; Heimpel and de Boer 2007). More recently, hymenopteran genetics has experienced a significant advance with the recent completion of the genome sequence for the honey bee {Apis mellifera) (The Honeybee Genome Consortium 2006), soon to be complemented by the genome sequencing initiative for Nasonia species (Human Genome Sequencing Center 2008). The honey bee genome sequence is of major economic interest because of this species' importance as a pollinator, but at the same time it is of major scientific interest due to its role as a model organism for the study of eusociality. As such, its genome sequence is considered today as a key to better understanding the relationship between genomes and social behaviour (so-called "sociogenomics") (Robinson et al. 2005; Check 2006; The Honeybee Genome Consortium 2006; Wilson 2006). Similarly, the Nasonia genome sequences are expected to contribute significantly to the management of these parasitoids 92 as biological control agents, and at the same time, they are expected to provide new insights regarding what constitutes a parasitoid genetically and how that differs from eusocial genomes. Whether other properties of genomes in addition to sequence (e.g., genome size) relate to parasitism or sociality remains an open question. Genome size diversity Until very recently, only 10 genome size estimates were available for the entire order Hymenoptera (Gregory 2008) (Table 3.1), but fortunately this has begun to change. In particular, the list was recently expanded to include 40 new estimates for the family Formicidae (ants) (Tsutsui et al. 2008). Considering these data, hymenopteran genome sizes appeared to range from 0.16 pg in the parasitoid wasps Habrobracon juglandis and H. Serinopae to 0.77 pg in the fire ant Solenopsis invicta, with a mean value of 0.36 ± 0.02 pg. The new ant estimates provided by Tsutsui et al. (2008) make Hymenoptera the fourth best studied order among insects in the number of species sampled, but leaves all other groups within the order practically undescribed. Despite the lack of genome size data for most groups, it is apparent that hymenopterans have small genomes for insects: all of them are smaller than 1 pg and hence are well below the hypothetical 2 pg upper limit for holometabolous (complete metamorphosis) insects (Gregory 2002c, 2005; Tsutsui et al. 2008; Chapter 1). The small genomes in Hymenoptera, as in other holometabolous orders, may be the result of developmental constraints imposed by complete metamorphosis which consists of undergoing major changes in morphology through cell differentiation and growth in a short and limited amount of time during the pupal stage (Gregory 2002c). However, the 93 fact that their genomes are considerably below the hypothetical upper threshold for holometabolous insects suggests the existence of other restrictions which may be imposed by characteristics unique to this order, such as having a parasitoid or eusocial life- strategy (Gregory 2002c; Johnston et al. 2004; Koshikawa et al. 2008). Objectives This study aimed foremost to expand knowledge on genome size diversity in Hymenoptera by surveying some of the most familiar and economically important superfamilies: Apoidea, Chalcidoidea, Ichneumonoidea, and Vespoidea. Moreover, this survey also permitted a preliminary exploration of how bulk genome properties, and not just genome sequences (The Honeybee Genome Consortium 2006), may vary according to parasitism and eusociality. It has been suggested previously that parasites and/or parasitoids may exhibit small genomes due to developmental, physiological, or body size constraints (e.g., Johnston et al. 2004), and indeed a few parasitoid wasps possess genomes at the smallest end of the range known in insects (GS = 0.10 pg, family Braconidae; T. R Gregory, unpublished). However, this hypothesis has never been tested through a comparison of parasitoid and non-parasitoid species from related taxonomic groupings. Hence, this study sampled parasitoids from the superfamilies Chalcidoidea and Ichneumonoidea to examine whether parasitoid genomes generally are small when compared to non- parasitoid hymenopterans. Similarly, the relationship between genome size and eusociality in Hymenoptera has not been examined previously. The basis for a potential correlation comes from the 94 effect of genome size on neuron size which, when braincase size is held roughly constant, can exert effects on brain complexity (Roth et al. 1994; Gregory 2002c). The limited data available for eusocial and solitary hymenopteran species make predictions difficult, especially since most values belong to ants (Formicidae) which are all eusocial. Therefore, this study sampled eusocial and solitary species from the superfamilies Vespoidea and Apoidea to examine the general link, if any, between genome size and eusociality. Materials and Methods Collection, identification, and vouchering of specimens In this study, hymenopterans were hand-collected primarily in Guelph and surrounding area (ON, Canada) between June and September 2006 and 2007. Samples were also collected in Orlando (University of Central Florida, FL, United States) and in the Archbold Biological Station (Lake Placid, central Florida) during May 2007. Samples of the ant Atta texana were collected by Gary Umphrey (Brownsville, Texas). Parasitoid wasp samples were donated by Beneficial Insectary and Biobest Canada. Specimens were identified by Gary Umphrey (University of Guelph), Matthias Buck (University of Guelph), Jason Gibbs (York University, Ontario), Stuart Fullerton (University of Central Florida), and Mark Deyrup (Archbold Research Station). Ant vouchers were placed in 99% ethanol and kept at room temperature. Bees and wasps were pinned and stored at room temperature. Vouchers are stored in the Gregory lab at the University of Guelph. The possible relationship between genome size and parasitism was evaluated by comparing genome size means and ranges between the parasitoid superfamilies 95 Ichneumonoidea and Chalcidoidea and the non-parasitoid superfamilies Apoidea and Vespoidea. Similarly, the relationship between genome size and social strategy was evaluated by comparing ranges and mean values between solitary and eusocial taxonomic groups (genera, families) that are closely related within the superfamilies Apoidea and Vespoidea. Genome size estimation Genome size estimates for hymenopterans collected in Ontario were obtained using flow cytometry (FCM) (Dolezel et al 2005; Chapter 1). The tissue selected for this method was head tissue obtained from fresh individuals that were then co-prepared with one head from the standard Drosophila melanogaster Oregon R strain female (GS = 0.18pg; Rasch 1971). Estimates for insects collected in Florida were obtained using Feulgen image analysis densitometry (FIA) from haemocyte samples and using haemocytes from Tenebrio molitor (GS = 0.52pg; Juan and Petitpierre 1989) as the standard (Hardie et al. 2002; Chapter 1). Results Genome size diversity Genome size estimates were obtained for 87 hymenopteran species from 17 families belonging to the superfamilies Apoidea, Chalcidoidea, Ichneumonoidea, and Vespoidea (Table 3.2). The mean value for the suborder Apocrita was 0.38 ± 0.02 pg while values ranged from 0.10 pg (the parasitoid wasp Aphidius colemani) to 1.14 pg (the solitary wasp Sceliphron caementarium). Nested ANOVA analysis showed that 42.53% of the variance was found the superfamily level, 4.21% at the family level, 46.13% at genus level, and 7.13% at the species level. 96 The Apoidea displayed the highest mean value (GS mean = 0.54 ± 0.04, n = 27), followed by Chalcidoidea (GS mean = 0.44 ± 0.08, n = 8), Vespoidea (GS mean = 0.31 ± 0.02, n = 48), and Ichneumonoidea (GS mean = 0.17 ± 0.04, n = 4) (Figure 3.3). Ranges for Apoidea were roughly twice (GS range = 0.19-1.14 pg, n = 27) that of Vespoidea (GS range = 0.15-0.55 pg, n = 48) and average values for these sister aculeate superfamilies differed significantly whether data for ants (Vespoidea: Formicidae n = 30 species) were included (ANOVA, F,, 73 = 41.3,p < 0.0001) or excluded (ANOVA, Flt43 = 20.8,p < 0.001). In addition, Apoidea and Vespoidea average estimates were considerably larger than that of the Ichneumonoidea, which contained the smallest range (GS range = 0.10- 0.27 pg, n = 4) and some of the smallest values reported for the order (Figure 3.3). Lastly, the superfamily Chalcidoidea displayed an intermediate range (GS range = 0.18-0.75 pg, n = 8) when compared to the aculeate superfamilies sampled in this study. Genome size diversity amons hymenopteran parasitoids The braconids Aphidius colemani (GS = 0.10 pg) and A. ervi (GS = 0.14 pg) contained the smallest genomes reported for the order and some of the smallest genomes ever reported in insects. Values for the family Braconidae (Ichneumonoidea) showed the smallest range (GS range = 0.10 - 0.16 pg, n = 3) observed among all parasitoid families sampled in this study and did not overlap with the only estimate available (Neotherrinia bicincta, GS = 0.27 pg) for the only other family (Ichneumonidae) recognized for the order (Belshaw et al. 1998). The largest parasitoid genomes were found in the superfamily Chalcidoidea, in Eretmocerus mundus (GS = 0.75 pg) and Aphelinus abdominalis (GS = 0.65 pg), both belonging to the family Aphelinidae. This family displayed the highest values among 97 chalcidoids (GS range = 0.42 - 0.75 pg, n = 4) while considerably lower values were found in the families Eulophidae (0.23 pg, n = 1) and Trichogrammatidae (0.18-0.19 pg, n = 2). The only estimate obtained for the family Encyrtidae was similar in size to the values reported here for Aphelinidae {Diglaiphus isaea, 0.56 pg). Genome size diversity at the family level in the superfamilies Vespoidea and Apoidea The range and number of species for the four Vespoidea and seven Apoidea families sampled in this study are shown in Figure 3.4. The only families sampled in this study with eusocial species were Apidae and Halictidae from the superfamily Apoidea and Formicidae and Vespidae from the superfamily Vespoidea. Overall, there were no significant differences in genome size estimates within Aculeata superfamilies between solitary and eusocial taxa examined in this study. In Vespoidea, the only estimate obtained here for the solitary family Scoliidae {Campsomeris plumipes fossulana, GS = 0.22 pg) was within the range of values obtained for the sister taxon Vespidae (GS range = 0.15-0.48 pg, GS mean = 0.26, n = 14) (Figure 3.4). The majority of Vespidae estimates obtained in this study belonged to eusocial hymenopterans from the genera Polistes (GS range = 0.24-0.48 pg, n = 4, subfamily Polistinae), Vespula (GS range = 0.17-0.23 pg, n = 4, subfamily Vespinae), and Dolichovespula (GS range = 0.32 pg, n = 1, subfamily Vespinae) (Table 3.3). The two Vespinae genera are sister groups (Carpenter 1987). There were no significant differences (ANOVA, F\^ = 20.8,p - 0.0671) in average values between Polistinae and Vespinae species. The five remaining estimates belonged to solitary species from five different genera from the subfamily Eumeninae (GS range = 0.15-0.30 pg, n = 5) (Table 3.3). Values for the solitary subfamily Eumeninae did not differ significantly from the 98 eusocial and sister (Hines et al. 2007) subfamilies Polistinae (ANOVA, Fi,6 = 3.1 \,p = 0.1282) and Vespinae (ANOVA, Fi, 7 = 0.07,p = 0.7965). The eusocial family Formicidae displayed similar ranges (GS range = 0.18-0.55 pg, n = 30) and the average value did not differ significantly (ANOVA, Fj, 42 = 1.67, p = 0.2003) from average values obtained for Vespidae (eusocial and solitary) or Scoliidae (solitary). Lastly, the family Mutillidae was represented in this study by three solitary species from only one genus (Dasymutilla) which showed values (GS range = 0.46-0.49 pg, n = 3) within the maximum values obtained for the families Vespidae and Formicidae (Figure 3.4). In Apoidea, the solitary family Sphecidae showed the largest range (GS range = 0.19-1.14 pg, n = 6) reported for any family in the order (Figure 3.4). This range was considerably larger than that of the solitary family Crabronidae (GS range = 0.35-0.66 pg, n = 4) and the monophyletic group of "bees" (GS range = 0.24-0.90 pg, n = 4) (Figure 3.4). However, the mean genome size for the solitary family Sphecidae (GS mean = 0.52 ±0.13) did not differ significantly (ANOVA, Fu» = 0.08,p = 0.7804) from that of the family Crabronidae (GS mean = 0.47 ± 0.07) and from that of "bees" (GS mean = 0.57 ± 0.04). Overall, the seven "bee" families showed similar genome size ranges (see Figure 3.3). The lowest value was found in the eusocial honey bee A. mellifera (Apidae; GS = 0.24 pg) and the highest value was found in the solitary bee Augochloropsis metallica (Halictidae; GS = 0.90 pg). Average estimates between the long-tongued families Apidae (eusocial and solitary) and Megachilidae (solitary) did not differ significantly (ANOVA, F\t9 = 3.72,p = 0.0854) nor did average values for the short-tongued families Andrenidae (solitary), Colletidae (solitary), Halictidae (eusocial and solitary) (ANOVA, 99 F2,3 = 0.19, p = 0.8395). There was no statistical difference between long-tongued and short-tongued families (ANOVA, F, 3 = 0.0l,p = 0.9448). In the family Apidae, estimates for solitary bees from the genus Ceratina were larger (GS range - 0.59-0.68 pg, n = 2) than that of the eusocial Apidae species belonging to the genera Bombus and Apis (GS range = 0.24-0.37 pg, n = 3) but the difference was marginal (ANOVA, Fi, 3 = 8.31,;? = 0.0634) (Table 3.3). Similar results were obtained for solitary (GS range = 0.57-0.67 pg, n = 2) and eusocial species (GS range = 0.40-0.60 pg, n = 2) of the family Halictidae (ANOVA, Fi>3 = 4.65,;? = 0.1639) (Table 3.3). Discussion Remarks on methods Estimates obtained using FCM in this study did not differ significantly from previous studies using the same method (Table 3.4). However, in this study FCM estimates were typically higher than FIA estimates by about 0.1 pg. FCM estimates were obtained from samples collected in southern Ontario and were determined from head tissue and using D. melanogaster as the standard, while all FIA samples came from Florida and were measured from haemocytes and using T. molitor as the standard. In one case (the ant Pseudomyrmex gracilis) no differences were found between different methodologies and populations. However, the mean estimates for the wasp Sceliphron caementarium differed considerably between individuals collected in southern Ontario (GS = 0.81 pg) and Florida (GS = 1.14 pg). This may be caused by differences in methodology, by other factors such as the presence of substances found in these wasps 100 that interfered with or enhanced staining (see Chapter 1), or may be suggestive of a cryptic species. A similar discrepancy was observed between genome estimates obtained in this and previous studies for the fire ant Solenopsis invicta. The first estimate obtained for this species used reassociation kinetics (GS = 0.77 pg, Li and Heinz 2000) and FCM (GS = 0.62 pg, Johnston et al. 2004) and differed considerably from the FIA estimate obtained in this study (GS = 0.48 pg). Such discrepancies may also be due to the different methodologies used between these studies. However, the S. invicta estimate obtained here is found within the ranges reported for the other two ant species described from the same genus: S. molesta (GS = 0.38 pg; this study using FCM) and S. xiloni (GS = 0.48; Tsutsui et al. 2008 using FCM). Hymenoptera genome size diversity The 87 hymenopteran genome size estimates provided in this study along with the 40 Formicidae estimates recently reported by Tsutsui et al. (2008) position the order Hymenoptera as the second best (number of species) characterized in insects (see Table 1.1). Compared to all other insect orders studied so far (with the exception of the Strepsiptera, for which only two estimates are available; Johnston et al. 2004), hymenopteran genomes exhibited the smallest average value, the smallest range, and some of the smallest genomes found in insects. Thus, not only is their distribution below the hypothetical 2pg threshold for holometabolous orders confirmed, but it suggests that additional factors are responsible for limiting genome size diversity in these insects. 101 Genome size in parasitoids Based on the results obtained in this study it may be concluded that parasitoid hymenopterans do not always have exceptionally small genomes when compared to non- parasitoids. Instead, the results suggest that exceptionally small genomes are characteristic of the family Braconidae. This is also suggested by the fact that the estimate (GS = 0.27 pg) obtained for another parasitoid species in the same superfamily is roughly twice that of any braconid value. In addition, this value is larger or similar in size to nearly half of the estimates available for the non-parasitoid species of Aculeata, which are contained in the sister clade of the Ichneumonoidea. Furthermore, most parasitoids from the superfamily Chalcidodea displayed intermediate to large genomes for the suborder, supporting the view that parasitoid hymenopterans are not restricted to extremely small values within Hymenoptera. Genome size vs eusociality Genome size did not correlate with eusociality in the hymenopteran groups examined in this study, rejecting any obvious link between genome size and social complexity in Hymenoptera. At first glance, a link may have become apparent since all available estimates for eusocial hymenopterans are below 0.60 pg, however, the majority of solitary species possessed values below this threshold as well. These conclusions disagree with the findings by Koshikawa et al. (2008) who suggested a link between eusociality and small genomes in the sister orders Blattaria (cockroaches) and Isoptera (termites). They reported that eusocial termites contained smaller genome size values that did not overlap with those of solitary cockroaches. 102 However, conclusions from this study are highly limited by having a small sample size, and more importantly, by comparing two insect orders which, although sharing a common origin, differ also in many other ways in their biology, ecology, and behaviour, and which may also explain the differences observed in genome size between these two groups. Evidently, the size of the genome does not relate to the evolution of social complexity in hymenopterans. Instead, genome features other than its size seem to have a predominant role, such as with the presence of "social" genes (The Honeybee Genome Consortium 2006) and their expression (Robinson et al. 2005). Future studies This study has served to explore for the first time the extent of genome size variation and its relationship to two of the most characteristic features for Hymenoptera: parasitism and eusociality. However, this should be seen as a preliminary survey only, and future genome size studies in this order may serve to examine in more detail how these features vary in this order. Further evaluation of the relation between genome size and parasitism may be performed in the hymenopteran superfamily Cynipoidea. This is a relatively small group (roughly 3,000 species) containing five families in which four are parasitoids and one (Cynipidae) contains non-parasitoid phytophagous species (gall wasps) (Grimaldi and Engel 2005). In addition to this, the classification and phylogeny for the superfamily Cynipoidea has been reviewed recently (Ronquist 1995, 1999; Vardal et al. 2003) which 103 would allow a closer examination of the evolution of these traits in a phylogenetic context. Although this study did not find evidence to support a link between genome size and eusociality, this relationship may be further examined in groups displaying all levels of social complexity. The monophyletic group of "bees" may be ideal for this since they have solitary species and eusocial species with different degrees of complexity (see Michener 2000). However, the main limitation in this group as in all other hymenopterans is that social behaviour for most bee species remains unknown (Dr. Laurence Packer, personal communications). In addition, future research should also aim to provide estimates for other superfamilies not reported in this study to test whether most hymenopteran genome sizes are below 1 pg. More importantly, they may also aim to identify other traits (e.g. body size, developmental rate, and flight) that may better describe the variation observed within this order and examine why hymenopteran genomes are on average roughly half the size of those observed in other holometabolous orders. 104 Table 3.1. Existing hymenopteran genome size estimates (GS, in pg) from the literature prior to the recent study of 40 ant species by Tsutsui et al. (2008) and the present analysis. Method abbreviations: RK) Reassociation kinetics; FD) Feulgen densitometry, FCM) Flow cytometry. References are provided at the end of the table. Taxonomy GS (pg) Method Reference Suborder Apocrita Superfamily Apoidea Family Apidae (Honey bees, bumble bees, carpenter bees, and their relatives) Apis cerana 0.19 RK 4 Apis mellifera 0.27 FCM 5 Bombus terrestris 0.42 FCM 6 Family Megachilidae (Leafcutter bees, mason bees, and their relatives) Megachile rotundata 0.30 RK 4 Superfamily Chalcidoidea Family Pteromalidae Nasonia vitripens 0.34 FIAD 1 Family Trichogrammatidae Trichogramma brassiccae 0.25 FCM 2 105 Taxonomy GS (pg) Method Reference Superfamily Ichneumonoidea Family Braconidae (Braconid wasps) Habrobracon juglandis 0.16 FIAD 1 Habrobracon serinopae 0.16 FIAD 3 Superfamily Vespoidea Family Formicidae (Ants) Solenopsis invicta 0.62,0.77 RK, FCM 2,7 Family Vespidae Polistes dominulus 0.31 FCM 2 References: 1) Rasch et al. (1975), 2) Johnston et al. (2004), 3) Rasch et al. (1977) 4) Jordan and Brosemer (1974), 5) The Honeybee Genome Consortium (2006), 6) Gadau et al (2001), 7) Li and Heinz (2000). 106 Table 3.2. Genome size estimates (GS, in pg) for 87 hymenopteran species including the standard error (± SE), number of individuals (N) including male (M) and female (F), and collection site (L). Information about lifestyle (LS) is provided for each species: parasitoid (PA), solitary (S), and eusocial (EU). Flow cytometry (FCM) and Feulgen image analysis densitometry (FIA) estimates were obtained using the standards Drosophila melanogaster Oregon R strain (GS = 0.18 pg; Rasch et al. 1971) and Tenebrio molitor (GS = 0.52 pg; Juan and Petitpierre 1989) respectively. Higher level classification follows Marshall (2007). Keys for collection site are provided at the end of the table. Taxonomy GS(pg) + /-SE N Method LS Suborder Apocrita 107 Superfamily Apoidea (bees and related solitary wasps) Family Andrenidae (Adrenid bees) Andrena dunningi 0.50 0.009 1M, 3F FCM Family Apidae (Honey bees, bumble bees, carpenter bees, and their relatives) Apis mellifera 0.24 0.004 5F FCM EU Bombus bimaculatus 0.34 0.008 3F FCM EU Taxonomy GS(pg) + /-SE N Method L LS Bombus impatiens 0.47 0.027 4F FCM 1 EU Ceratina calcarata 0.68 0.008 2M, 2F FCM 1 S Ceratina dupla dupla 0.59 IF FCM 1 . S Mellisodes desponsa 0.52 IF FCM 1 unknown Mellisodes iliata 0.37 0.007 3F FCM 1 unknown 108 Xylocopa virginica krombeini 0.69 IF FIA 4 unknown Family Crabronidae (Digger wasps and relatives) IF FCM Ectemnius continuus 0.38 IF FCM Gorytes atricornis 0.48 Microbembex monodonta 0.66 IF FIA Trypoxylon politum 0.35 2F FIA Taxonomy GS(pg) + /-SE N Method LS Family Colletidae (Plasterer bees and yellow faced bees) Haylaeus affinis 0.64 4F FCM Family Halictidae (Sweet bees and their relatives) Agapostemon splendens 0.66 6M,4F FIA Augochloropsis metallica 0.90 IF FCM 109 Halictus ligatus 0.49 0.011 8F FIA EU Halictus ligatus 0.60 0.017 6F FCM EU Halictus poeyi 0.40 0.001 3F FIA EU Family Megachilidae (Leafcutter bees, mason bees, and their relatives) Anthidiellum notatum rufimacuatum 0.48 IF FIA 4 unknown Megachile albitarsis 0.80 IF FIA 4 unknown Taxonomy + /-SE N Method L LS Megachile rotundata 0.83 IF FCM 1 S Family Sphecidae (Thread-waisted wasps) Ammophila pictipennis 0.41 IF FIA 5 S Chalybion californicum 0.54 - ' 4M, 2F FCM 1 S Larrabicolor 0.19 0.005 3F FCM 1 S Miscophus slossonae 0.19 0.002 2F FIA 2 S Sceliphron caementarium 0.81 0.009 2F FIA 5 S Sceliphron caementarium 1.14 0.045 1M, IF FCM 1 S Stictiella serrata 0.78 IF FIA 2 S Superfamily Chalcidoidea Family Aphelinidae Taxonomy GS(pg) + /-SE N Method L LS Aphelinus abdominalis 0.65 0.002 7F FCM 6 PA Encarsia formosa 0.42 0.002 7F FCM 6 PA Eretmocerus eremicus 0.55 0.008 7M, IF FCM 6 PA Eretmocerus mundus 0.75 0.008 5M, 2F FCM 6 PA Family Encyrtidae Leptomastix dactylopii 0.56 0.008 6F FCM 6 PA Family Eulophidae Diglyphus isaea 0.23 0.007 5M, IF FCM 6 PA Family Trichogrammatidae Trichogramma platneri 0.18* 0.007 3M, 7F FCM 7 PA Trichogramma pretiosum 0.19* 0.007 1M, 10F FCM 7 PA Taxonomy + /-SE N Method L LS Superfamily Ichneumonoidea Family Braconidae (Braconid wasps) Aphidius colemani 0.10 0.001 4M, 8F FCM 6 PA Aphidius ervi 0.14 0.002 5M, 2F FCM 6 PA Dacnusa sibirica 0.16 0.001 3M, 3F FCM 6 PA Family Ichneumonidae (Ichneumonid wasps) Neotheronia bicincta 0.27 0.006 3F FIA 5 PA Superfamily Vespoidea Family Formicidae (Ants) Amblyopone pallipes 0.37 IF FCM 1 EU Dolichoderus taschenbergi 0.23 0.008 - 3F FCM 1 EU Taxonomy GS(pg) + /- SE N Method L LS Dolichoderus mariae 0.18f 3F FCM 1 EU Dorymyrmex bureni 0.18 0.003 3F FIA 2 EU Forelius pruinosus 0.22 0.003 2F FIA 2 EU Tapinoma sessile 0.37 0.007 10F FCM 1 EU Camponotus floridanus 0.23 0.002 3F FIA 2 EU Lasius (Acanthomyops) latipes 0.27 0.004 2M, 9F FCM 1 EU Lasius minutus 0.23 0.001 4F FCM 1 EU Parathechina longicornis 0.18 0.005 3F FIA 2 EU Aphenogaster (rudis-texana, "N 16") 0.43 IF FCM 1 EU Aphenogaster (rudis-texana, "N .17") 0.46 0.011 4F FCM 1 EU Taxonomy GS(pg) + /-SE N Method L LS Aphenogaster (rudis-texana, "N 22b") 0.44 0.001 3F FCM EU Aphenogaster fulva 0.42 0.002 3F FCM 1 EU Aphenogaster treatae 0.50 0.019 3F FCM EU Atta texana 0.27 0.009 6F FCM EU Monomorium viride 0.50 0.019 3F FIA EU 114 Pheidole dentata 0.24 0.003 3F FIA EU Pheidole floridana 0.21 0.004 5F FIA EU Solenopsis invicta 0.47 0.005 4F FIA EU Solenopsis molesta 0.38 0.004 5F FCM 1 EU Tetramorium caespitum 0.27 0.009 3F FCM EU Temnothorax ambiguus 0.31 0.003 9F FCM EU Taxonomy GS(pg) + /-SE N Method L LS Temnothorax texanus 0.32 0.017 5F FCM 1 EU Trachymyrmex septentrionalis 0.25 0.002 3M,6F FIA 2 EU Odontomachus brunneus 0.33 0.014 3F FIA 2 EU Ponera pennsylvanica 0.55 0.014 6F FCM 1 EU Pseudomyrmex ejectus 0.29 IF FIA 2 EU Pseudomyrmex gracilis 0.35 5F FCM, FIA 1,2 EU Family Mutillidae (Velvet ants) Dasymutilla archboldi 0.46 0.015 2F FIA Dasymutilla phyrrus 0.49 IF FIA Dasymutilla occidentalis 0.46 IF FIA Family Scoliidae (Scoliid wasps) Taxonomy GS(pg) + /-SE N_ Method L LS Campsomeris plumipes fossulana 0.21 0.004 4F FIAD 4 S Family Vespidae Dolichovespula arenaria 0.32 IF FCM 1 EU Eumenes fraternus 0.22 IF FIA 5 S Eumenes smithii 0.30 IF FIA 2 S Euodynerus cf hidalgo 0.30 IF FIA 4 S Polistes dominula 0.29 0.004 6F FCM 1 EU Polistes dorsalis 0.48 0.009 7F FIA 2,4,5 EU Polistes fuscatus 0.31 0.019 2F FIA 5 EU Polistes fuscatus 0.41 0.006 4F FCM 1 EU Polistes metricus 0.24 0.005 4F FIA 2,4,5 EU Taxonomy GS (pg) + /- SE N Method L LS Symmorphus canadensis 0.23 IF FCM Vespula bulgaris 0.18* IF FCM EU Vespula germanica 0.23 0.003 1M,3F FCM EU Vespula maculifrons 0.22 0.003 4F FCM EU Vespula squamosa 0.17 0.005 3F FIA EU Zethus slossonae 0.15 IF FIA Location (L): 1) Guelph (ON), 2) Archbold Research Station (Lake Placid, FL), 3) Brownsville (TX), 4) Orlando (FL), 5) Kissimmee (FL), 6) Biobest Canada, 7) Beneficial Insectary Canada. *FCM estimates in which unknown 2C peak overlapped with standard and this was confirmed by running an unknown sample without the standard. fFCM estimate in which the head of the unknown was cut in half, one half was co-prepared with the standard, and the other half was measured by itself. Table 3.3. Genome size ranges and means (GS, in pg), and species number (N) for eusocial and solitary genera and families that are closely related in Aculeata. Information about lifestyle (LS) is provided for each species: solitary (S), and eusocial (EU). Taxonomy GS Range GS Mean N LS Superfamily Apoidea Family Apidae Genus Apis 0.27 0.27 1 EU 118 0.37-0.47 0.42 2 EU Genus Bombus 0.57-0.67 0.62 2 S Genus Ceratina Family Halictidae 0.9.0 0.90 1 S Genus Agapostemon 0.66 0.66 2 S Genus Augochloropsis 0.40-0.55 0.48 2 EU Genus Halictus Taxonomy GS Range GS Mean N Superfamily Vespoidea Family Vespidae Genus Dolichovespula 0.27 0.27 1 EU Genus Vespula 0.17-0.23 0.22 2 EU Genus Polistes 0.37 - 0.47 0.42 2 EU 119 Genus Euodynerus 0.30 0.30 1 S Genus Eumenes 0.22 - 0.30 0.25 2 S Genus Symmorphus 0.23 0.23 1 S Genus Zethus 0.15 0.15 1 S Table 3.4. Genome size estimates (GS, in pg) for six ant species obtained in this study compared to estimates reported by Tsutsui et al. (2008). Methods are specified as follows: flow cytometry (FCM) and Feulgen image analysis densitometry (FIA). Estimates from the two studies were highly correlated (r = 0.90, p = 0.015, n = 6) and there were no significant differences among them overall (paired t-test, p = 0.6). Species This study Tsutsui et al (2008) GS Method GS Method Amblyopone pallipes 0.37 FCM 0.34 FCM 120 Odontomachus brunneus 0.33 FIA 0.44 FCM Ponera pennsylvanica 0.55 FCM 0.60 FCM Pseudomyrmex gracilis 0.35 FIA 0.40 FCM Tapinoma sessile 0.37 FCM 0.38 FCM Tetramorium caespitum 0.27 FCM 0.26 FCM Stephanoidea Trigonalyoidea Megalyroidea E m m | vanioidea Ichneumonoidea <: Chrysidoidea Apoidea <=^ Vsspoidoa <"—i Platygastroidea Ceraphronoidea Mymmarommatoidea Chalcidoidea <=^ Cynipoidea Proctotrupoidea Figure 3.1. Phylogenetic relationships of the Apocrita superfamilies modified from Grimaldi and Engel (2005) by showing the three Aculeata superfamilies (Chrysidoidea, Apoidea, and Vespoidea) and their phylogenetic relationships. Arrows show the superfamilies sampled in this study. 121 Plumariidae Scolebythrdae Bethyltdae Chrysididat Chrysidoidea Sclerogibbidae DryWdae Embolemidae SierolomorpWidae Rhopalosomatidae Bradynobaenidae Formicidae Vesptdae Vespoidea Scoidae Pompilidae Mutiliidae Sapygidae TlpWidae Hetero^r»aeidae Ampulicidae Sphecidae 1 Apoidea Crabronidae Bees* Figure 3.2. The phytogeny of "Aculeata" (Chrysidoidea, Vespoidea, and Apoidea) after Brothers and Carpenter (1993), and modified from Grimaldi and Engel (2005). Grey branches show the only hymenopteran taxa with eusocial species. * "Bees" is a monophyletic group composed of the families Andrenidae, Apidae, Colletidae, Halictidae, Megachilidae, Melittidae, and Steronotritidae (Michener 2000). 122 1.20 j 1.00 0,80 ' I GO CO 0.60 0 0.40 4 0.20 I ApoWea Chalcidoidea* tchrseumonoldea* Vespoidea SuperfamiJy Figure 3.3. Genome size (GS) ranges (black bars) and means (white lines) for the hymenopteran superfamilies sampled in this study. Numbers in brackets indicate the number of species analyzed per superfamily. * Parasitoid superfamilies. 123 Chryskioidea Slerolomorphidae Rhopalosomatidat Bradynobaenidae Forrnicidae <6s«««e*o.i8-o.5s Pg, n= 30 Vespidae |<3S range - 0.15-0.48 pg, n = 14) Vespoidea SCOidae m range * 0.22 pg, n = 1) Pompiiidae Mutillidae grange * 0.46-0.49 pg, n - 3) Sapygidae Tiphiidae Heteragynaeidat Ampuiicidae Sphecidae «» »njc= 0.19-1.14 Pg, n= e> Crabronidae (6$ range- 0.35-0.66 Pg, n= 4) Apidae {6$ range* 0.24-0.69 pg, n = 8) Apoidea Megachilidae Andrenldae m range* 0.49 Pg, n= D Figure 3.4. Genome size ranges and number of species measured (n) for the Vespoidea and Apoidea families sampled in this study. The aculeate phylogeny (after Brothers and Carpenter 1993 as presented by Grimaldi and Engel 2005) was modified by including all bee families for which phylogenetic relationships remain inconclusive (Michener 2000). Grey branches show the only hymenopteran taxa with eusocial species. 124 Conclusion The 187 odonate and hymenopteran genome size estimates provided in this study increased the current insect dataset by -40%, demonstrating the feasibility of greatly expanding the information available for insects using the modern methods of flow cytometry (FCM) and Feulgen image analysis densitometry (FIA). In addition to providing possible new target organisms for whole-genome sequencing projects, the data obtained provided a new opportunity to tackle some of the most interesting questions in genome evolution research including the relationships between genome size and body size, development, flight, a parasitoid lifestyle, and social complexity. In odonates, genome sizes displayed a moderate range (0.41-2.38 pg) but exceeded the hypothetical 2 pg upper limit as expected for hemimetabolous insects (Gregory 2002). Genome size did not correlate with voltinism, nymphal habitat, or body size. Interestingly, genome size varied in relation to flight strategy, where flier odonates possessed larger genomes than perchers. This suggested for the first time in insects a possible link between genome size and flight-related metabolic rate similar to that observed in birds. Hymenopterans possessed on average the smallest genomes reported in any reasonably well-studied insect order and were found to be considerably below the 2 pg upper limited expected for holometabolous insects. Parasitoid hymenopteran genome sizes displayed similar ranges to that observed among non-parasitoids suggesting that evolving a parasitoid life strategy is not necessarily linked to having exceptionally small 125 genomes. Similarly, genome size did not differ significantly between solitary and eusocial hymenopterans. The results reported in this study strongly support the view that genome size does not vary randomly in insects, and that its variation is correlated with other important aspects of insect biology and behaviour. Application of a similar approach in other insect groups in future studies would certainly provide important new insights to better understand the evolution of the genome and the causes, effects, and mechanisms that regulate its size and content. 126 References Adams, M.D. et al. 2000. The genome sequence of Drosophila melanogaster. Science 287:2185-2195. Abro, A. 1999. The size range of sperm bundles in aeshnid dragonflies (Anisoptera: Aeshnidae). Odonatologica 28:273-277. Abro, A. 2000. Sperm clusters in Zygoptera (Coenagrionidae, Lestidae, Calopterygidae). Odonatologica 29:51-56. Agopian, S. and Mola, L. 1988. Intra and interspecific karyotype variability in five species of Libellulidae (Anisoptera, Odonata). Caryologia 41:69-78. Bennett, M. D. 1971. The duration of meiosis. Proceedings of the Royal Society of London Series B-Biological Sciences 178:277-299. Bennett, M. D. 1972. Nuclear DNA content and minimum generation time in herbaceous plants. Proceedings of the Royal Society of London Series B-Biological Sciences 181:109-135. Bennett, M.D., Leitch, I.J., Price, H.J., and Johnston, J.S. 2003. Comparisons with Caenorhabditis (-100 Mb) and Drosophila (-175 Mb) using flow cytometry show genome size in Arabidopsis to be -157 Mb and thus -25% larger than the Arabidopsis Genome Initiative estimate of-125 Mb. Annals of Botany 91: 547- 557. Bried, J. T., Bennett, L. W., and Ervin, G. N. 2005. Live mass and length-mass allometry of adult odonates collected in east-central Mississippi, United States. Odonatologica 34:111-122. 127 Bried, J. T. and Ervin, G. N. 2007. Intraspecific models and spatiotemporal context of size - mass relationships in adult dragonflies. Journal of the North American Benthological Society 26:681-693. Brothers, D.J. 1999. Phylogeny and evolution of wasps, ants and bees (Hymenoptera, Chrysoidea, Vespoidea, and Apoidea). Zoological Scripta 28: 233-249. Brothers, D.J. and Carpenter, R.M. (1993). Phylogeny of the Aculeata: Chrysidoidea and Vespoidea (Hymenoptera). Journal of Hymenoptera Research 2: 227-304. Burton, D.W., Bickham, J.W., and Genoways, H.H. 1989. Flow-cytometric analyses of nuclear DNA content in four families of neotropical bats. Evolution 43: 756-765. Carpenter, J.M. 1987. Phylogenetic relationships and classification of the Vespinae (Hymenoptera: Vespidae). Systematic Entomology 12:413-431. Cavalier-Smith, T. 1978. Nuclear volume control by nucleoskeletal DNA, selection for cell volume and cell growth rate, and the solution to the DNA C-value paradox. Journal of Cell Science 34: 247-278. Cavalier-Smith, T. 1982. Skeletal DNA and the evolution of genome size. Annual Review of Biophysics and Bioengineering 34: 247-278. Check, E. 2006. From hive minds to humans. Nature 448: 893. Corbet, P. S. 1962. A biology of dragonflies. H. F. & G. Witherby Ltd, London, UK. Corbet, P. S. 1999. Dragonflies: behaviour and ecology of Odonata. Cornell University Press, New York, USA. Corbet, P. S., F. Suhling, and D. Soendgerath. 2006. Voltinism of Odonata: a review. International Journal of Odonatology 9:1 -44. 128 Cruden, R. W. 1968. Chromosome numbers of some North American dragonflies (Odonata). Canadian Journal of Genetics and Cytology 10:200-214. Cumming, R. 1964. Cytogenetic studies in the order Odonata. PhD Thesis. University of Texas, Austin, USA. Dasgupta, J. 1957. Cytological studies on the Indian dragonflies II. A study of the chromosomes during meiosis in thirty species of Indian Odonata (Insecta). Proceedings of the Zoological Society Calcutta 10: 1—66. DeSalle, R., Gregory, T.R., and Johnston, J.S. 2005. Preparation of samples for comparative studies of arthropod chromosomes: Visualization, in situ hybridization, and genome size estimation. Methods in Enzymology 395: 460- 488. Dole el, J., Bartos, J., Voglmayr, H., and Greilhuber, J. 2003. Nuclear DNA content and genome size of trout and human. Cytometry Part A 51 A: 127-128. Dole el, J., Greilhuber, J., and Suda, J. 2007. Flow cytometry with plant cells: analysis of genes, chromosomes, and genomes. Wiley-VCH, Weinheim, Germany. Ferrari, J.A. and Rai, K.S. 1989. Phenotypic correlates of genome size variation in Aedes albopinctus. Evolution 43: 895-899. Ferreira, A., Kiauta, B., and Zaha, A. 1979. Male germ cell chromosomes of thirty-two Brazilian dragonflies. Odonatologica 8: 5-22. Finston, T.L., Hebert, P.D.N., and Foottit, R.B. 1995. Genome size variation in aphids. Insect Biochemistry and Molecular Biology 25: 189-196. 129 Forbes, M.R., Schalk, G., Miller, J.G., and Richardson, J.M.L. 1997. Male-female morph interactions in the damselfly Nehalennia irene (Hagen). Canadian Journal of Zoology 75:253-260. Gadau, J., Gerloffs, C.U., Krugers, N., Chan, H., Schmid-Hempel, P., Wille, A., and Page, R.E. 2001. A linkage analysis of sex determination in Bombus terrestris (L.) (Hymenoptera: Apidae). Heredity 87: 234-242. Grabow, K. and Ruppell, G. 1995. Wing loading in relation to size and flight characteristics of European Odonata. Odonatologica 24:175-186. Gregory, T.R., Hebert, P.D.N., and Kolasa, J. 2000. Evolutionary implications of the relationship between genome size and body size in flatworms and copepods. Heredity 84:201-20. Gregory, T. R. 2001. Coincidence, coevolution, or causation? DNA content, cell size, and the C-value enigma. Biological Reviews 76:65-101. Gregory, T. R. 2002a. A bird's-eye view of the C-value enigma: Genome size, cell size, and metabolic rate in the class Aves. Evolution 56:121-130. Gregory, T. R. 2002b. Genome size and developmental complexity. Genetica 115:131- 146. Gregory, T. R. 2002c. Genome size and developmental parameters in the homeothermic vertebrates. Genome 45:833-838. Gregory, T.R. and Hebert, P.D.N. 2003. Genome size variation in lepidopteran insects. Canadian Journal of Zoology 81: 1399-1405. Gregory, T.R. and Johnston, J.S. 2008. Genome size diversity in the family Drosophilidae. Heredity 101: 228-238. 130 Gregory, T. R., Nedved, O., and Adamowicz, S.J. 2003. C-value estimates for 31 species of ladybird beetles (Coleoptera: Coccinellidae). Hereditas 139: 121-127. Gregory, T. R. 2005. Genome size evolution in animals. In The Evolution of the Genome. Edited by T.R. Gregory. Elsevier, San Diego, USA, 3-87. Gregory, T. R. 2008. Animal Genome Size Database. Available online: http ://www. genomesize.com. Greilhuber, J., Dole el, J., Lysak, M.A., and Bennet, M.D. 2005. The origin, evolution and proposed stabilization of the terms 'genome size' and 'C-value' to describe nuclear DNA contents. Annals of Botany 95:255-260. Grimaldi, D. and Engel, D.S. 2005. Evolution of Insects. Cambridge University Press. New York, USA. Hardie, D. C, Gregory, T. R., and Hebert, P. D. N. 2002. From pixels to picograms: A beginners'guide to genome quantification by Feulgen image analysis densitometry. Journal of Histochemistry & Cytochemistry 50:735-749. Hebert, P.D.N., Cywinska, A., Ball, S.L., and deWaard, J.R. 2003. Biological identifications through DNA barcodes. Proceedings of the Royal Society of London B 270:313-321. Heimpel, G.E. and de Boer, J.G. 2007. Sex determination in Hymenoptera. Annual Reviews of Entomology 53: 209-230. Heinrich, B. 1974. Thermoregulation in endothermic insects. Science 185:747-756. Heinrich, B. and Casey, T. 1978. Heat transfer in dragonflies: 'fliers' and 'perchers'. Journal of Experimental Biology 74:17-36. 131 Hillier et al. 2004. Sequence and comparative analysis of the chicken genome provide unique perspectives on vertebrate evolution. Nature 432:695-716. Hines, H.M., Hunt, J.H., O'Connor, T.K., Gillespie, J.J., Cameron, S.A. 2007. Proceedings of the National Academy of Science of the United States of America 104: 3295-3299. Holldobler, B. and Wilson, E.O. 1990. The Ants. Havard University Press., Cambridge, USA. Hughes, A.L. 1999. Adaptive Evolution of Genes and Genomes. Oxford University Press., Oxford, UK. Hughes, A.L. and Hughes, M.K. 1995. Small genomes for better flyers. Nature 377:391- 391. Human Genome Sequencing Center. 2008. The Nasonia genome project. Available online: http://www.hgsc.bcm.tmc.edu/projects/nasonia Johnston, T.A., and Cunjak, R.A. 1999. Dry mass-length relationships forbenthic insects: a review with new data from Catamaran Brook, New Brunswick, Canada. Freshwater Biology 41:653-674. Johnston, J.S., Ross, L.D., Beani, L., Hughes, D.P., and Kathirithamby, J. 2004. Tiny genomes and endoreduplication in Strepsiptera. Insect Molecular Biology 13(6): 581-585. Jordan, R.A. and Brosemer, R.W. 1974. Characterization of DNA from three bee species. Journal of Insect Physiology 20: 2513-2520. 132 Josephson, R.K., Malamud, J. G. and Stokes, D. R. 2000. Asynchronous muscle: A primer. Journal of Experimental Biology 203:2713-2722. Juan, C. and Petitpierre, E. 1989. C-banding and DNA content in seven species of Tenebrionidae (Coleoptera). Genome 32:834-839. Kaunelas, M.S. 1970. Factors affecting ribosome isolation from adult crickets, Acheta domesticus. Journal of Insect Physiology 16: 813-824. Kiauta, B. 1967. Considerations on the evolution of the chromosome complement in Odonata. Genetica 38:430-466. Kiauta, B. 1968. Variation in size of the dragonfly m-chromosome, with considerations on its significance for the choreography and taxonomy of the order Odonata, and notes on the validity of the rule of Reinig. Genetica 39:64--1'A. Kiauta, B. 1969a. Sex chromosomes and sex determining mechanisms in Odonata, with a review of the cytological conditions in the family Gomphiidae, and references to the karyotypic evolution in the order. Genetica 40:127-157. Kiauta, B. 1969b. Autosomal fragmentations and fusions in Odonata and their evolutionary implications. Genetica 40:158-180. Kiauta, B. 1972. Synopsis of the main taxonomic data in the order Odonata. Odonatologica 1:73-102. Kiauta, B. and Van Brink, J.M. 1978. Male chromosome complements of some Florida dragonflies, United States. Odonatologica 7: 15-25. Koshikawa, S., Miyazaki, S., Cornette, R., Matsumoto, T., and Miura, T. 2008. Genome size of termites (Insecta, Dictyoptera, Isoptera) and wood roaches (Insecta, Dyctyoptera, and Crytocercidae). Naturwissenschaften 95: 859-867. 133 Kurnick, N.B. and Herskowitz, I.H. 1952. The estimation of polyteny on Drosophila salivary gland nuclei based on determination of desoxyribonucleic acid content. Journal of Cellular Comparative Physiology 39: 281-299. Li, J. and Heinz, K.M. 2000. Genome complexity and organization in the red imported fire ant Solenopsis invicta Buren. Genetical Research 75: 129-135. Mangan, K.F. and Kaunelas, M.S. 1970. Nuclease activity of ribosomal and soluble fractions of Acheta domesticus. Journal of Insect Physiology 16: 1155-1163. Marden, J. H. 1987. Maximum lift production during takeoff in flying animals. Journal of Experimental Biology 130:235-258. Marshall, S. E. 2007. Sawflies, wasps, bees, and ants. In Insects: Their Natural History and diversity. Firefly Books Ltd, Richmond Hill, Canada, 518-584. May, M. L. 1976. Thermoregulation and adaptation to temperature in dragonflies (Odonata: Anisoptera). Ecological Monographs 46:1-32. May, M. L. 1981. Allometric analysis of body and wing dimensions of male Anisoptera. Odonatologica 10:279-292. May, M. L. 1991. Dragonfly flight-power requirements at high speed and acceleration. Journal of Experimental Biology 158:325-342. Michener, CD. 2000. The Bees of the World. The John Hopkins University Press. Baltimore, USA. Misof, B., Rickert, A.M., Buckley, T.R. Fleck, G., and Sauer, K.P. 2001. Phylogenetic signal and its decay in mitochrondrial SSU and LSU rRNA gene fragments of Anisoptera. Molecular Biology and Evolution 18: 27-37. 134 Mola, L. M. and A. G. Papeschi. 1994. Karyotype evolution in Aeshna (Aeshnidae, Odonata). Hereditas 121:185-189. Needham, J. D., Westfall, M. J., and May, M. L. 2000. Dragon/lies of North America. Revised edition. Scientific Publishers, Gainesville, USA. Oguma, K. 1930. A comparative study of the spermatocyte chromosome in allied species of the dragonfly. Journal of the Faculty of Science, Hokkaido University. Series 6, Zoology: 1:1—32. Olmo, E. 1983. Nucleotype and cell size in vertebrates: a review. Basic and Applied Histochemistry 27: 227-256. Organ, C.L., Shedlock, A.M., Meade, A., Pagel, M., Edwards, S.V. 2007. Origin of avian genome size and structure in non-avian dinosaurs. Nature 446: 180-184 Page, E.R.J., Gadau, J., and Beye, M. 2002. The emergence of hymenopteran insects. Genetics 160: 375-379. Perepelov, E., Bugrov, A.G., and Warchalowska-Sliwa, E. 1998. C-banded karyotypes of some dragonfly species from Russia. Folia Biologica-Krakow 46:137-142. Perepelov, E., Bugrov, A.G., and Warchalowska-Sliwa, E. 2001. C-banded Karyotypes of some dragonfly species from Russia. II. The families Cordulegasteridae, Corduliidae and Gomphidae. Folia Biologica-Krakow 49:175-178. Petrov, N.B. and Aleshin, V.V. 1983. Heterogeneity and homology of the repetitive and unique DNA-sequences of dragonflies (Odonata, Insecta). Molecular Biology 17:277-286. 135 Petrov, N.B., Aljeshin, V.V., and Antonov, A.S. 1984. DNA sequence organization in 2 odonate species Calopteryx splendens (Harris) and Aeshna coerulea (Strom) (Zygoptera: Calopterygidae; Anisoptera: Aeshnidae) Odonatologica 13:269-279. Prasad, R. and Thomas, K.I. 1992. C-band pattern homogenity in dragonflies (Odonata). Caryologia 45:57-68. Rasch, E.M. 1985. DNA "standards" and the range of accurate DNA estimates by Feulgen absorption microspectrophotometry. In Advances in Microscopy. Edited by R.R. Cowden and S.H. Harrison. Alan R. Liss, New York. Pages 137-166. Rasch, E.M., Barr, H.J., and R.W. Rasch. 1971. DNA content of sperm of Drosophila melanogaster. Chromosoma 33:1-18. Rasch, E.M., Cassidy, J.D., and King, R.C. 1975. Somatic cell DNA content in haploid, diploid, and triploid adults of Mormoniella vitripennis. Journal of Cell Biology 67: 353A Rasch, E.M., Cassidy, J.D., and King, R.C. 1977. Evidence for dosage compensation in parthenogenetic hymenoptera. Chromosoma 59: 323-340 Ray-Chaudhuri, S.P. and Dasgupta, J. 1949. Cytological studies on the Indian dragonflies (Odonata). Proceedings of the Zoological Society Calcutta 2:81-93. Richter, M.R. 2000. Social wasps (Hymenoptera: Vespidae) foraging behaviour. Annual Reviews in Entomology 45: 121-150. Robinson, G.E., Grozinger, CM., and Whitfield, C.W. 2005. Sociogenomics: social life in molecular terms. Nature Reviews Genetics 6: 257-270. 136 Ronquist, F. 1995. Phylogeny and early evolution of the Cynipoidea (Hymenoptera). Systematic Entomology 46: 1-74. Ronquist, F. 1999. Phylogeny, classification, and evolution of the Cynipoidea (Hymenoptera). Zoological Scripta 28: 139-164. Roth, G., J. Blanke, and D.B. Wake (1994). Cell size predicts morphological complexity in the brains of frogs and salamanders. Proceedings of the National Academy of Sciences of the United States of America 91: 4796-4800 Rowe, L. and Ludwig, D. 1991. Size and timing of metamorphosis in complex life-cycles - time constraints and variation. Ecology 72:413-427. Riippell, G. 1989. Kinematic analysis of symmetrical flight manoeuvres of Odonata. Journal of Experimental Biology 144:13—42. Schorr, M., Lindeboom, M., and Paulson, D. 2008. World Odonata list. University of Puget Sound, Tacoma, USA. Available online: http://www.ups.edu/x6140.xml Seshachar, B.R. and Bagga S. 1962. Chromosome number and sex-determining mechanism in the dragonfly Hemianax ephippiger (Burmeister). Cytologia 27:443^149. Seshachar, B.R. and Bagga S. 1963. Cytology and cytochemistry of spermiogenesis in two dragonflies Ictinogomphus rapax (Rambur) and Pantalaflavescens. Journal ofMorphology 113:267-213. Sformo, T., Doak, P. 2006. Thermal ecology of interior Alaska dragonflies (Odonata: Anisoptera). Functional Ecology 20:114-123. 137 Simmons, P. 1977a. Neuronal control of dragonfly flight 1: Anatomy. Journal of Experimental Biology 71:123-140. Simmons, P. 1977b. Neuronal control of dragonfly flight 1: Physiology. Journal of Experimental Biology 71:141-155. Smith, D. S. 1960. Innervation of the fibrillar flight muscle of an insect: Tenebrio molitor (Coleoptera). Journal of Biochemical and Biophysical Cytology 8:447-466. Smith, D. S. 1961a. The organization of the flight muscle in a dragondly, Aeshna sp. (Odonata). Journal of Cell Biology 11:119-145. Smith, D. S. 1961b. The structure of insect fibrillar flight muscle: A study made with special reference to the membrane systems of the fiber. Journal of Biochemical and Biophysical Cytology 10:123-158. Smith, D. S. 1966. The organization of flight muscle fibers in the Odonata. Journal of Cell Biology 28:109-126. Smith, E.A. 1916. Spermatogenesis of the dragonfly Sympetrum semicinctum (Say) with remarks upon Libellula basalis. The Biological Bulletin 31: 269-302. Szarski, H. 1970. Changes in the amount of DNA in cell nuclei during vertebrate evolution. Nature 226: 651-652 Szarski, H. 1983. Cell size and the concept of wasteful and frugal evolutionary strategies. Journal of Theoretical Biology 105: 201-209. Thomas, A.L.R., Taylor, G.K., Srygley, R.B., Nudds, R.L., and Bomphrey, R.J. 2004. Dragonfly flight: free-flight snd tethred flow visualizations reveal a diverse array of unsteady lift-generating mechanisms,controlled primarily via angle of attack. The Journal of Experimental Biology 207: 4499^4-323. 138 Tiersch, T.R. and Wachtel, S.S. 1991. On the evolution of genome size in birds. Journal of Heredity 82: 363-368 Tsutsui, N.D., Suarez, A.V., Spagna, J.C., and Johnston, J.S. 2008. The evolution of genome size in ants. Evolutionary Biology 8:64 doi:10.1186/1471-2148-8-64 The Honeybee Genome Consortium. 2006. Insights into social insects from the genome of the honeybee Apis mellifera. Nature 443: 931-949 Van Brink, J.M. and Kiauta, B. 1964. Notes on chromosome behaviour in the spermatogenesis of the damselfly Enallagma cyathigerum (Charp.) (Odonata: Coenagrionidae). Genetica 35: 171-174. Van Gossum, H., K., Beirinckx, M., Forbes, R., and Sherratt, T. N. 2007. Do current hypotheses explain continental and seasonal variation in female morph frequencies of the damselfly, Nehalennia irenel Biological Journal of the Linnean Society 90:501-508. Vardal, H., Sahlen, G., and Ronquist, F. 2003. Morphology and evolution of the cynipoid egg (Hymenoptera). Zoological Journal of the Linnean Society 139: 247-260. Vilhar, B., J. Greilhuber, J.D. Koce, E.M. Temsch, and Dermastia, M. 2001. Plant genome size measurement with DNA image cytometry. Annals of Botany 87: 719-728. Wakeling, J. M. and Ellington, C. P. 1997. Dragonfly flight 2: Velocities, accelerations and kinematics of flapping flight. Journal of Experimental Biology 200:557-582. Ware, J., May, M., and Kjer, K. 2007. Phylogeny of the higher Libelluloidea (Anisoptera: Odonata): an exploration of the most speciose superfamily of dragonflies. Molecular Phylogenetics and Evolution 45: 289—310. 139 Westerman, M., Barton, N.H., and Hewitt, G.M. 1987. Differences in DNA content between two chromosomal races of the grasshopper Podisma pedestris. Heredity 58:221-228. Westfall, M. J. and May, M. L. 1996. Damsel/lies of North America. Scientific Publishers, Gainsville, USA. Wilmore, P.J. and Brown, A.K. 1975. Molecular properties of orthopteran DNA. Chromosoma 51: 337-345. Wilson, E.O. 2006. How to make a social insect. Nature 443: 919-920 140 Appendix 1: Body size measurements with photographic image analysis (PIA) This appendix provides detailed descriptions for the setup of PIA for Odonata. Pictures were taken by placing specimens on white papered on top of a cooler flash box (Picture 1). Most specimens were air dried in cellophane envelopes with their wings folded back. Therefore, pictures were taken with the specimens laying sideways on a flat surface (Picture 2) The few dragonflies that dried with their wings unfolded were place on the flat surface such that pictures were taken from their dorsal side (Picture 3). There was no need in either case to separate the wings by relaxing or by cutting them from the specimens to make wing length and area measurements. Camera Settings Mode Av orM ISO 100 Shutter Speed 1/250 F stop 16 Meter Type Evaluative Lens Type 100 mm MACRO Exposure Compensation Yes +/- 2 stops White Balance flash Color Temperature 5200 Flash types Flash Settings Ratio / Power Values MT-24EX Manual Al/4 : Bl/4 : Cl/1 CHANNEL #1 Each with: Each facing one click slipcovers/white satin on inward of 180° 2-430 EX Manual 1/8 Power CHANNEL #1 Both facing inside white Styrofoam lab cooler Paper used 1 sheet regular copy 1 sheet Vellum on top 2-430 EX Manual 1/2 Power CHANNEL #1 141 Copy-Stand: • Copy camera setup: Polaroid MP-4 Land Camera Image Resolution: • 3504x2336x24b Imaging Software: • NIS Element Imaging Software, Basic Research • NIS-Elements BR 2.30, SP4, Hotfix (Build 389). Copyright 1991-2007. Laboratory Imaging, Nikon. • Applications Used: Measure Length and Area in pixels. Computer Hardware: • Intel (R) Core (TM)2 CPU T7200 @ 2.00 GHz. 2.00 GHz, 0.98 GB of RAM. Computer System: • Microsoft Windows XP. Professional. Version 2002. Service Pack 2. Computer Screen: • iMAC (Apple) Widescren 20 inch monitor. • Screen resolution: 1440 by 900 pixels. • Colour Quality (32 bit) 142 Picture 1. PIA setup for Odonata. Details are provided in the text. 143 Picture 2. Sample picture of the damselfiy Ischnura ramburii from its side view. A 10 mm2 millimetre paper between two glass cover slips (25 x 25 mm) held together with 2 small pieces of tape was placed as the scale along with the specimen ID. Image dimensions were 3504x2336x24bi (JPEG format). 144 Picture 3. Sample picture of the dragonfly Gomphus (Hylogomphus) geminatus from its dorsal view. A 10 mm millimetre paper between two glass cover slips (25 x 25 mm) held together with 2 small pieces of tape was placed as the scale along with the specimen ID. Image dimensions were 3504x2336x24bi (JPEG format). 145 Appendix 2. Pearson correlations for odonate genome size and body size means. Variables include, genome size (GS), minimum body weight (TW), body length (BL), head length (HL), thorax length (TH), abdomen length (AL), forewing length (FWL), hindwing length (HWL), forewing area (FWA), and hindwing area (HWA). All mean values are in Table 2.8. Calculations were performed with species that have measurements for all variables only. Bolded p-values are not siginificant (p > 0.05). Odonata species Correlations (Pearson ) GS TW TL HL THL AL FWL HWL FWA TW 0..088 6 P-VALUE 0..388 0 TL 0..239 8 0..880 3 0..018 0 <0..000 1 HL 0..085 3 0.,940 4 0.,775 1 0..406 3 <0..000 1 <0..000 1 THL -0..006 3 0..969 5 0..863 9 0..918 0 0..951 1 <0..000 1 <0..000 1 <0..000 1 AL 0..317 1 0,.736 1 0.,964 6 0..597 1 0.7057 147 0..001 6 <0..000 1 <0.,000 1 <0..000 1 <0.0001 FWL 0..120 8 0..943 6 0..909 1 0..916 2 0.9339 0.7905 0..238 7 <0..000 1 <0..000 1 <0,.000 1 <0.0001 <0.0001 HWL 0..125 8 0..945 7 0..910 1 0..919 3 0.9341 0.7910 0.9988 0..219 4 <0..000 1 <0..000 1 <0..000 1 <0.0001 <0.0001 <0.0001 FWA 0..070 9 0..946 3 0..851 8 0..940 2 0.9313 0.7078 0.9788 0.9815 0..489 9 <0..000 1 <0..000 1 <0..000 1 <0.0001 <0.0001 <0.0001 <0.0001 HWA 0..051 3 0,.944 7 0..815 8 0,.954 9 0.9303 0.6561 0.9653 0.9686 0.9948 0..618 1 <0..000 1 <0..000 1 <0,.000 1 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 Cases Included 97 Odonata genera Correlations (Pearson) GS TW TL HL THL AL FWL HWL FWA TW 0..153 8 P-VALUE 0..343 4 TL 0,.318 8 0,.868 9 0..045 0 <0..000 1 HL 0,.153 0 0..905 1 0..768 4 0..345 9 <0..000 1 <0..000 1 THL 0..029 9 0..954 3 0..848 3 0..897 5 0,.854 5 <0..000 1 <0..000 1 <0..000 1 148 AL 0..384 0 0..753 3 0..974 2 0..622 0 0.7139 0..014 4 <0..000 1 <0..000 1 <0..000 1 <0.0001 FWL 0..151 7 0..911 2 0..911 7 0..906 3 0.9086 0.8205 0..350 0 <0..000 1 <0..000 1 <0..000 1 <0.0001 <0.0001 HWL 0..164 0 0..915 1 0..912 8 0..909 9 0.9095 0.8206 0.9989 0..311 9 <0..000 1 <0..000 1 <0..000 1 <0.0001 <0.0001 <0.0001 FWA 0..103 1 0..921 8 0..858 0 0..933 4 0.9169 0.7431 0.9833 0.9855 0..526 6 <0..000 1 <0..000 1 <0..000 1 <0.0001 <0.0001 <0.0001 <0.0001 HWA 0..062 7 0..915 9 0..808 6 0,.949 1 0.9123 0.6782 0.9630 0.9658 0.9932 0..700 9 <0..000 1 <0..000 1 <0..000 1 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 Cases Included 4 0 Odonata Familiy Correlations (Pearson) GS TW TL HL THL AL FWL HWL FWA TW 0.2284 P-VALUE 0.5545 TL 0.2900 0.9207 0.4490 0.0004 HL 0.3199 0.9716 0.8465 0.4013 <0.0001 0.0040 THL 0.1199 0.9868 0.9171 0.9419 0.7586 <0.0001 0.0005 0.0001 AL 0.3183 0.8126 0.9741 0.7129 0.8077 0.4039 0.0078 0.0000 0.0311 0.0085 FWL 0.2819 0.9721 0.9407 0.9545 0.9606 0.8495 0.4623 <0.0001 0.0002 0.0001 0.0000 0.0037 HWL 0.2907 0.9705 0.9420 0.9544 0.9563 0.8527 0.9997 0.4479 <0.0001 0.0001 0.0001 0.0001 0.0035 <0.0001 FWA 0.2819 0.9613 0.8968 0.9542 0.9411 0.7923 0.9897 0.9894 0.4624 <0.0001 0.0011 0.0001 0.0002 0.0109 <0.0001 <0.0001 HWA 0.2730 0.9663 0.8664 0.9682 0.94 36 0.7455 0.9792 0.9785 0.9951 0.4772 <0.0001 0.0025 <0.0001 0.0001 0.0211 <0.0001 <0.0001 <0.0001 Cases Included 9 Anisoptera (dragonflies) species Statistix 8.0 Correlations (Pearson) GS TW TL HL THL AL FWL HWL FWA TW 0.4998 P-VALUE <0.0001 TL 0.5313 0.9294 <0.0001 <0.0001 HL 0.6218 0.7701 0.8006 <0.0001 <0.0001 <0.0001 THL 0.3034 0.8913 0.8771 0.6723 0.0175 <0.0001 <0.0001 <0.0001 AL 0.5397 0.9018 0.9929 0.7653 0.8232 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 FWL 0.4746 0.8667 0.9035 0.8523 0.8352 0.8759 0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 HWL 0.5002 0.8720 0.9082 0.8580 0.8336 0.8811 0.9978 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 FWA 0.4976 0.8563 0.8851 0.8454 0.8128 0.8573 0.9902 0.9911 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 HWA 0.4844 0.8276 0.8541 0.8401 0.7802 0.8260 0.9828 0.9849 0.9941 0.0001 <0.0001 <0.0001 <0.0001 <0.0001 •<0.0001 <0.0001 <0.0001 <0.0001 Cases Included 61 Anisoptera genera Correlations (Pearson) GS TW TL HL THL AL FWL HWL FWA TW 0..486 1 P-VALUE 0..004 1 TL 0..504 3 0..907 1 0..002 8 <0..000 1 HL 0..617 5 0..762 8 0.,835 8 0..000 1 <0..000 1 <0.,000 1 THL 0,.300 3 0..889 1 0..885 9 0,.721 6 0,.089 5 <0..000 1 <0..000 1 <0..000 1 151 AL 0..507 9 0..879 2 0.,992 6 0..798 4 0.8344 0,.002 5 <0..000 1 <0..000 1 <0,.000 1 <0.0001 FWL 0..410 2 0..838 8 0.,921 4 0..884 3 0.8393 0.8971 0..017 7 <0..000 1 <0.,000 1 <0..000 1 <0.0001 <0.0001 HWL 0,.432 6 0..846 8 0..924 3 0,.892 5 0.8407 0.8990 0.9979 .0..011 9 <0..000 1 <0.,000 1 <0..000 1 <0.0001 <0.0001 <0.0001 FWA 0..418 2 0..842 4 0..906 2 0,.885 2 0.8311 0.8778 0.9922 0.9947 0,.015 4 <0,.000 1 <0..000 1 <0..000 1 <0.0001 <0.00Ol <0.0001 <0.0001 HWA 0..393 4 0..812 6 0..872 0 0,.876 1 0.7968 0.8428 0.9810 0.9849 0.9942 0..023 5 <0..000 1 <0..000 1 <0..000 1 <0.0001 <0. 0-001 <0.0001 <0.0001 <0.0001 Cases Included 33 Anisoptera family Correlations (Pearson) GS TW TL HL THL AL FWL HWL FWA TW 0.4522 P-VALUE- 0.3679 TL 0.4647 0..991 6 0.3531 0,.000 1 HL 0.7876 0..751 5 0..734 3 0.0629 0..085 0 0..096 5 THL 0.1354 0,.906 2 0,.922 5 0..545 8 0.7982 0..012 8 • 0..008 8 0..262 6 AL 0.4759 0..988 6 0,.996 8 0..708 9 0.9027 152 0.3400 0,.000 2 <0..000 1 0..114 8 0.0137 FWL 0.4845 0..929 7 0,.920 5 0..879 3 0.8578 0.8922 0.3301 0,.007 2 0..009 2 0..021 0 0.0289 0.0168 HWL 0.5045 0..897 8 0..867 8 0..906 4 0.7757 0.8401 0.9836 0.3075 0,.015 1 0,.025 1 0..012 7 0.0698 0.0363 0.0004 FWA 0.5234 0,.928 4 0..918 9 0..895 5 0.8360 0.8928 0.9985 0.9860 0.2866 0,.007 5 0,.009 6 0..015 8 0.0382 0.0166 <0.0001 0.0003 HWA 0.5142 0,.856 6 0,.824 1 0..905 6 0.7144 0.7964 0.9644 0.9920 0.9711 0.2967 0..029 4 0..043 7 0..012 9 0.1107 0.0580 0.0019 0.0001 0.0012 Cases Included 6 Zygoptera (damselflies) species Correlations (Pearson) GS TW TL HL THL AL FWL HWL FWA TW -0..525 7 P-VALUE 0..001 0 TL -0..479 4 0..929 9 0..003 1 <0..000 1 HL -0..501 5 0..946 7 0..877 3 0,.001 8 <0,.000 1 <0..000 1 THL -0..548 4 0..968 4 0..959 0 0..904 6 0..000 5 <0..000 1 <0..000 1 <0..000 1 AL -P..454 1 0..902 3 0..996 9 0..845 5 0..935 5 153 0..005 4 <0..000 1 <0..000 1 <0..000 1 <0..000 1 FWL -0..374 1 0..942 1 0..954 1 0..905 1 0..928 5 0..943 7 0..024 6 <0..000 1 <0..000 1 <0..000 1 <0..000 1 <0..000 1 HWL -0..398 5 0..939 4 0..957 7 0..906 0 0..924 9 0..948 9 0..997 1 0,.016 1 <0..000 1 •<0..000 1 <0..000 1 <0..000 1 <0..000 1 <0..000 1 FWA -0..374 2 0..887 1 0..889 5 0..859 7 0..836 2 0..884 7 0..946 9 0.,956 5 0,.024 6 <0..000 1 <0..000 1 <0..000 1 <0..000 1 <0..000 1 <0..000 1 <0..000 1 - HWA -0,.351 7 0,.881 4 0..891 4 0..850 8 0..830 9 0..888 8 0..947 3 0,.957 4 0.9972 0..035 4 <0..000 1 <0..000 1 <0..000 1 <0..000 1 <0..000 1 <0..000 1 <0,.000 1 <0.0001 Cases Included 36 All Zygoptera genera Correlations (Pearson) GS TW TL HL THL AL FWL HWL FWA TW -0.5309 P-VALUE 0.2201 TL -0.4710 0.9912 0.2860 <0.0001 HL -0.5235 0.9957 0.9933 0.2279 <0.0001 <0.0001 THL -0.5770 0.9976 0.9869 0.9921 0.1750 <0.0001 <0.0001 <0.0001 AL -0.4450 0.9860 0.9993 0.9894 0.9804 0.3171 <0.0001 <0.0001 <0.0001 0.0001 FWL -0.3828 0.9730 0.9866 0.9838 0.9604 0.9882 0.3967 0.0002 <0.0001 0.0001 0.00,06 <0.0001 HWL -0.3875 0.9720 0.9882 0.9833 0.9601 0.9903 0.9995 0.3905 0.0002 <0.0001 0.0001 0.0006 <0.0001 <0.0001 FWA -0.3179 0.9441 0.9701 0.9590 0.9258 0.9758 0.9880 0.9903 0.4872 0.0014 0.0003 0.0006 0.0028 0.0002 <0.0001 <0.0001 HWA -0.3033 0.9348 0.9616 0.9505 0.9146 0.9679 0.9826 0.9852 0.9993 0.5085 0.0020 0-.0005 0.0010 0.0039 0.0003 0.0001 0.0001 <0.0001 Cases Included 7 All Zygoptera family Correlations (Pearson) GS TW TL HL THL AL FWL HWL FWA TW -0.3164 P-VALUE 0.7951 TL -0.2793 0.9992 0.8198 0.0248 HL -0.2763 0.9991 1.0000 0.8218 0.0267 0.0020 THL -0.4188 0.9939 0.9889 0.9884 0.7249 0.0701 0.0949 0.0969 AL -0.2521 0.9977 0.9996 0.9997 0.9843 0.8378 0.0427 0.0180 0.0160 0.1129 FWL -0.0930 0.9739 0.9820 0.9826 0.9431 0.9870 0.9407 0.1457 0.1209 0.1189 0.2158 0.1029 HWL -0.1100 0.9777 0.9851 0.9856 0.9486 0.9896 0.9999 0.9298 0.1348 0.1100 0.1081 0.2049 0.0921 0.0109 FWA 0.0371 0.9362 0.9492 0.9502 0.8919 0.9577 0.9915 0.9892 0.9764 0.2286 0.2038 0.2018 0.2987 0.1858 0.0829 0.0938 HWA 0.0629 0.9269 0.9408 0.9418 0.8799 0.9499 0.9878 0.9851 0.9997 0.9599 0.2450 0.2202 0.2183 0.3152 0.2023 0.0993 0.1102 0.0164 Cases Included 3